//connect to visionary
private void button_Connect_Click(object sender, EventArgs e)
{
String host = textBox_IP.Text;
int port = 2114;
VisionaryDataStream dataStream = new VisionaryDataStream(host, port);
VisionaryControl control = new VisionaryControl(host);
textBox_Index.Text = "0";
Task.Run(async() =>
{
// Buffer for reading data
Byte[] bytes = new Byte[1024];
try
{
IPAddress ipAddress = Dns.Resolve("localhost").AddressList[0];
TcpListener server = new TcpListener(ipAddress, Convert.ToInt32(textBox_BackendPort.Text));
System.Threading.Thread.Sleep(1000);
server.Start();
//MessageBox.Show("Waiting for client to connect...");
TcpClient client = server.AcceptTcpClient();
NetworkStream nStream = client.GetStream();
//network loop, send then wait for reply loop
int i = -1;
while (true)
{
Thread.Sleep(2);
i = nStream.Read(bytes, 0, bytes.Length);
if (i > 0)
{
String data = null;
// Translate data bytes to a ASCII string.
data = System.Text.Encoding.ASCII.GetString(bytes, 0, i);
//MessageBox.Show("Received:" + data);
// Process the data sent by the client.
data = data.ToUpper();
//nStream.Write(bytes, 0, bytes.Length);
if (data == "REQUEST\n")
{
//transmit 1 frame
//compile image from 640x512x3 -> 983040x1
try
{
//byte[] bStream = ImageToByte(bitmap_RGB);
nStream.Write(bitmap_arry, 0, bitmap_arry.Length);
}
catch
{
}
}
else if (data != "EMPTY\n" && checkBox_UseBackend.Checked == true)
{
try
{
//try parsing x/y/w/h if its not empty
parsedata(data);
}
catch
{
}
}
}
}
}
catch (Exception e1)
{
MessageBox.Show("SocketException: " + e1.Message);
}
});
try
{
Task.Run(async() =>
{
if (!await dataStream.ConnectAsync())
{
// Data stream connection failed
MessageBox.Show("error");
}
if (!await control.ConnectAsync())
{
// Data control (CoLaB) connection failed
}
await control.StartAcquisitionAsync();
});
}
catch (Exception ex)
{
MessageBox.Show("ERROR:" + ex.ToString());
throw;
}
System.Threading.Thread.Sleep(1500);
//-------------------------------------------------------------------------------------------------------
// Receiving & Image Processing thread
//-------------------------------------------------------------------------------------------------------
Task.Run(async() =>
{
while (true)
{
//-------------------------------------------------------------------------------------------------------
// Read Data from Visionary Datastream
//-------------------------------------------------------------------------------------------------------
VisionaryFrame frame = await dataStream.GetNextFrameAsync();
//System.Threading.Thread.Sleep(1000);
VisionarySDepthMapData depthMap = frame.GetData <VisionarySDepthMapData>();
// Important: When converting multiple frames, make sure to re-use the same converter as it will result in much better performance.
PointCloudConverter converter = new PointCloudConverter();
Vector3[] pointCloud = converter.Convert(depthMap);
CenterH = pointCloud[250 * 640 + 320].Z;
//read and set range of textboxs
setTextboxRange();
//Assign converted image
bitmap = depthMap.ZMap.ToBitmap(Zmap_DR, Zmap_Offset);
bitmap_RGB = depthMap.RgbaMap.ToBitmap();
bitmap_arry = depthMap.RgbaMap.Data.ToArray();
ZMap_arry = depthMap.ZMap.Data.ToArray();
this.label1.Text = bitmap.GetPixel(320, 250).R.ToString();
//-------------------------------------------------------------------------------------------------------
// Optional default image proccessing method (locate box)
//-------------------------------------------------------------------------------------------------------
if (checkBox_MinAreaRect.Checked == true)
{
Bitmap TempMap = bitmap;
if (RGBAsZmap == true)
{
TempMap = bitmap_RGB;
}
//init different images for different detection stages
Image <Bgr, byte> a = new Image <Bgr, byte>(TempMap);
Image <Gray, byte> b = new Image <Gray, byte>(a.Width, a.Height); //edge detection
Image <Gray, byte> c = new Image <Gray, byte>(a.Width, a.Height); //find contour
//set threshold
int Blue_threshold = 50; //0-255
int Green_threshold = 50; //0-255
int Red_threshold = 50; //0-255
if (RGBAsZmap == false)
{
a = ~a;
a = a.ThresholdBinary(new Bgr(Blue_threshold, Green_threshold, Red_threshold), new Bgr(255, 255, 255));
}
//Set ROI
a.ROI = new Rectangle(ROIx, ROIy, (int)(640 * ROIScale), (int)(512 * ROIScale));
//Find edges
int cannytherhold = 100;
CvInvoke.Canny(a, b, cannytherhold / 2, cannytherhold, 3, false);
//Enhance canny edges
Mat struct_element = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(3, 3), new Point(-1, -1));
CvInvoke.Dilate(b, b, struct_element, new Point(-1, -1), 1, BorderType.Default, new MCvScalar(255, 255, 255));
//Find contours
VectorOfVectorOfPoint con = new VectorOfVectorOfPoint();
CvInvoke.FindContours(b, con, c, RetrType.List, ChainApproxMethod.ChainApproxNone);
Point[][] con1 = con.ToArrayOfArray();
PointF[][] con2 = Array.ConvertAll <Point[], PointF[]>(con1, new Converter <Point[], PointF[]>(PointToPointF));
listBox_BoxList.Items.Clear();
for (int i = 0; i < con.Size; i++)
{
//Filter params
double tempArea = CvInvoke.ContourArea(con[i], true);
double tempArc = CvInvoke.ArcLength(con[i], true);
double tempScale = tempArea / Math.Pow(tempArc / 4, 2);
if (tempArea >= MinPixelArea && tempScale > RatioFilter)
{
RotatedRect rrec = CvInvoke.MinAreaRect(con2[i]); //g
//-------------------------------------------------------------------------------------------------------
// find box dimensions
//-------------------------------------------------------------------------------------------------------
//find box height
int boxHeight = -10000;
int tempX = 0; //X axis offset if point is NULL
int tempY = 0; //Y axis offset if point is NULL
while (boxHeight < 0 && tempX <= 50)
{
boxHeight = (int)Math.Round(((double)BackgroundH - (double)pointCloud[(ROIy + tempY + (int)rrec.Center.Y) * 640 + (ROIx + tempX + (int)rrec.Center.X)].Z) * (double)1000);
tempX++;
}
tempX--;//for angle correction
//apply sensor prespective angle correction
if (AngleCorr == true)
{
double boxCenterOffset = Math.Sqrt(Math.Pow(pointCloud[(ROIy + (int)rrec.Center.Y) * 640 + (ROIx + tempX + (int)rrec.Center.X)].X, 2) + Math.Pow(pointCloud[(ROIy + (int)rrec.Center.Y) * 640 + (ROIx + tempX + (int)rrec.Center.X)].Y, 2));
double boxCenterAngle = Math.Atan(boxCenterOffset * 1.1 / BackgroundH);
double heightMulti = 1 / Math.Cos(boxCenterAngle);
boxHeight = (int)((double)boxHeight * heightMulti);
}
//find dimension of 1 pixel from avg of X/Y axis box plane
tempX = 0; //unused here
tempY = 0; //unused here
double PixelScaleX = -10000;
double PixelScaleY = -10000;
double PixelScale = -10000;
PixelScaleX = (double)(pointCloud[(int)(ROIy + rrec.Center.Y) * 640 + (int)(ROIx + rrec.Center.X) - (15 + tempX)].X - (double)pointCloud[(int)(ROIy + rrec.Center.Y) * 640 + (int)(ROIx + rrec.Center.X) + (15 + tempX)].X) * 1000.0 / (30.0 + (tempX + tempX));
PixelScaleY = (double)(pointCloud[(int)(ROIy + rrec.Center.Y - (15 + tempY)) * 640 + (int)(ROIx + rrec.Center.X)].Y - (double)pointCloud[(int)(ROIy + rrec.Center.Y + (15 + tempY)) * 640 + (int)(ROIx + rrec.Center.X)].Y) * 1000.0 / (30.0 + (tempY + tempY));
PixelScale = (PixelScaleX + PixelScaleY) / 2;
int boxWidth = (int)(rrec.Size.Width * PixelScale);
int boxLength = (int)(rrec.Size.Height * PixelScale);
//Rounding result
boxLength = (int)(Math.Round((double)boxLength / Rounding, MidpointRounding.AwayFromZero) * Rounding);
boxWidth = (int)(Math.Round((double)boxWidth / Rounding, MidpointRounding.AwayFromZero) * Rounding);
boxHeight = (int)(Math.Round((double)boxHeight / Rounding, MidpointRounding.AwayFromZero) * Rounding);
double boxArea = ((double)boxLength / 10) * ((double)boxWidth / 10) * ((double)boxHeight / 10);//cm
//add box to listbox
listBox_BoxList.Items.Add("Box (Length: " + boxLength + "mm, Width: " + boxWidth + "mm, Height: " + boxHeight + "mm, Vol:" + boxArea + "cm^3)");
PointF[] pointfs = rrec.GetVertices();
for (int j = 0; j < pointfs.Length; j++)
{
CvInvoke.Line(a, new Point((int)pointfs[j].X, (int)pointfs[j].Y), new Point((int)pointfs[(j + 1) % 4].X, (int)pointfs[(j + 1) % 4].Y), new MCvScalar(0, 255, 0, 255), 4);
}
}
}
//save box list
if (SaveQueueList == true)
{
System.IO.StreamWriter SaveFile = new System.IO.StreamWriter(textBox_Savepath.Text + "/Output/Detection Result/" + DateTime.Now.ToString("yyyyMMdd_hhmmss") + ".txt");
foreach (var item in listBox_BoxList.Items)
{
SaveFile.WriteLine(item);
}
SaveFile.Close();
SaveQueueList = false;
}
//save pointcloud
if (SaveQueuePly == true)
{
await PointCloudPlyWriter.WriteFormatPLYAsync(textBox_Savepath.Text + "/Output/PointCloud/" + index.ToString() + ".ply", pointCloud, depthMap.RgbaMap, true);
}
/*
* //for displaying contours
* for (int i = 0; i < con.Size; i++)
* {
* CvInvoke.DrawContours(d, con, i, new MCvScalar(255, 255, 0, 255), 2);
* }
*/
this.pictureBox2.Image = a.ToBitmap();
this.pictureBox1.Image = bitmap_RGB;
}
else
{
this.pictureBox2.Image = bitmap_RGB;
this.pictureBox1.Image = bitmap;
}
try
{
bitmap_Mixed = mixedMap(bitmap_arry, ZMap_arry, Convert.ToUInt16(textBox_DynamicRange.Text));
this.pictureBox_Mixed.Image = bitmap_Mixed;
}
catch { }
}
});
}
public void DisplayWebcam(popup popup)
{
while (true)
{
Mat frame = ReadCamera1();
int largestCircle = 0;
List <Shape> foundShape = new List <Shape>();
List <char> shapeType = new List <char>();
// resize to PictureBox aspect ratio
int newHeight = (frame.Size.Height * pictureBox0.Size.Width) / frame.Size.Width;
Size newSize = new Size(pictureBox0.Size.Width, newHeight);
CvInvoke.Resize(frame, frame, newSize);
Image <Hsv, Byte> img = frame.ToImage <Hsv, Byte>();
Image <Gray, byte> coralNestImg = frame.ToImage <Gray, Byte>();
Image <Gray, byte> coralOutcropImg = frame.ToImage <Gray, Byte>();
Image <Gray, byte> starfishImg = frame.ToImage <Gray, Byte>();
Image <Bgr, byte> drawnImage = frame.ToImage <Bgr, byte>();
Image <Bgr, byte> coralNestDrawn = frame.ToImage <Bgr, byte>();
Image <Bgr, byte> coralOutcropDrawn = frame.ToImage <Bgr, byte>();
Image <Bgr, byte> starfishDrawn = frame.ToImage <Bgr, byte>();
Image <Gray, byte> gridImg = frame.ToImage <Gray, Byte>();
Image <Bgr, byte> gridDrawn = frame.ToImage <Bgr, byte>();
//Line Follow
Hsv T_Lower = new Hsv(colorThresholds[4].h_Lower, colorThresholds[4].s_Lower, colorThresholds[4].v_Lower);
Hsv T_Upper = new Hsv(colorThresholds[4].h_Upper, colorThresholds[4].s_Upper, colorThresholds[4].v_Upper);
Image <Gray, byte> laneImg = frame.ToImage <Gray, Byte>();
laneImg = img.InRange(T_Lower, T_Upper).Erode(2).Dilate(2);
//count for columbs
int leftC = 0;
int centerC = 0;
int rightC = 0;
for (int i = (laneImg.Height / 20) * 15; i < laneImg.Height; i++)
{
for (int j = (laneImg.Width / 20) * 14; j < (laneImg.Width / 20) * 16; j++)
{
//left
if (laneImg.Data[i, j, 0] == 255)
{
leftC++;
}
}
for (int j = (laneImg.Width / 20) * 16; j < (laneImg.Width / 20) * 18; j++)
{
//center
if (laneImg.Data[i, j, 0] == 255)
{
centerC++;
}
}
for (int j = (laneImg.Width / 20) * 18; j < (laneImg.Width / 20) * 20; j++)
{
//right
if (laneImg.Data[i, j, 0] == 255)
{
rightC++;
}
}
}
if (leftC > centerC && leftC > rightC)
{
//twist left
Twist twist = new Twist();
twist.Linear.Z = 1f;
SetInput(twist);
label7.Invoke(new Action(() => {
label7.Text = "left";
}));
}
else
{
if (centerC > rightC && centerC > leftC)
{
// twist center
Twist twist = new Twist();
twist.Linear.X = 1f;
SetInput(twist);
label7.Invoke(new Action(() => {
label7.Text = "center";
}));
}
else
{
//twist right
Twist twist = new Twist();
twist.Linear.Z = -1f;
SetInput(twist);
label7.Invoke(new Action(() => {
label7.Text = "right";
}));
}
}
if (leftC + rightC + centerC <= 100)
{
Twist twist = new Twist();
twist.Linear.X = 0;
SetInput(twist);
//Update keys
popup.UpdateKeys(key);
}
// Find Centers and Draw Bounding Rectangle
for (int i = 0; i < 3; i++)
{
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint();
List <Shape> shapes = new List <Shape>();
Mat draw = new Mat();
Hsv t_Lower = new Hsv(colorThresholds[i].h_Lower, colorThresholds[i].s_Lower, colorThresholds[i].v_Lower);
Hsv t_Upper = new Hsv(colorThresholds[i].h_Upper, colorThresholds[i].s_Upper, colorThresholds[i].v_Upper);
int shapeIndex = 0;
double largestArea = 0;
switch (i)
{
case 0:
coralNestImg = img.InRange(t_Lower, t_Upper).Erode(2).Dilate(2);
CvInvoke.FindContours(coralNestImg, contours, draw, Emgu.CV.CvEnum.RetrType.Ccomp, ChainApproxMethod.ChainApproxSimple);
break;
case 1:
coralOutcropImg = img.InRange(t_Lower, t_Upper).Erode(2).Dilate(2);
CvInvoke.FindContours(coralOutcropImg, contours, draw, Emgu.CV.CvEnum.RetrType.Ccomp, ChainApproxMethod.ChainApproxSimple);
break;
case 2:
starfishImg = img.InRange(t_Lower, t_Upper).Erode(2).Dilate(2);
CvInvoke.FindContours(starfishImg, contours, draw, Emgu.CV.CvEnum.RetrType.Ccomp, ChainApproxMethod.ChainApproxSimple);
break;
default:
break;
}
for (int j = 0; j < contours.Size; j++)
{
double peramiter = CvInvoke.ArcLength(contours[j], true);
VectorOfPoint positions = new VectorOfPoint();
CvInvoke.ApproxPolyDP(contours[j], positions, 0.03 * peramiter, true);
var moments = CvInvoke.Moments((contours[j]));
shapes.Add(new Shape());
//Find Centroid of shape and area and store in shape class
shapes[j].x_Center = (moments.M10 / moments.M00);
shapes[j].y_Center = (moments.M01 / moments.M00);
shapes[j].area = CvInvoke.ContourArea(contours[j]);
}
//Only compute if there are shapes on screen
if (shapes.Count > 0)
{
// find largest shape sstore it
for (int j = 0; j < shapes.Count; j++)
{
if (shapes[j].area > largestArea)
{
largestArea = shapes[j].area;
shapeIndex = j;
}
}
Rectangle boundingRect = new Rectangle((int)(shapes[shapeIndex].x_Center - 50), (int)(shapes[shapeIndex].y_Center - 50), 100, 100);
switch (i)
{
case 0:
coralNestDrawn = coralNestImg.Convert <Bgr, byte>();
coralNestDrawn.Draw(boundingRect, new Bgr(0, 0, 255), 3, LineType.EightConnected, 0);
//Add to list of shapes found
foundShape.Add(shapes[shapeIndex]);
shapeType.Add('n');
break;
case 1:
coralOutcropDrawn = coralOutcropImg.Convert <Bgr, byte>();
coralOutcropDrawn.Draw(boundingRect, new Bgr(0, 0, 255), 3, LineType.EightConnected, 0);
//Add to list of shapes found
foundShape.Add(shapes[shapeIndex]);
shapeType.Add('o');
break;
case 2:
starfishDrawn = starfishImg.Convert <Bgr, byte>();
starfishDrawn.Draw(boundingRect, new Bgr(0, 0, 255), 3, LineType.EightConnected, 0);
//Add to list of shapes found
foundShape.Add(shapes[shapeIndex]);
shapeType.Add('s');
break;
}
}
}
//Circle Detection
/* // var findEdges = img.SmoothGaussian((int)(colorThresholds[3].h_Upper)).Convert<Gray, byte>().ThresholdBinaryInv(new Gray(colorThresholds[3].s_Lower), new Gray(colorThresholds[3].s_Upper)).Erode((int)(colorThresholds[3].h_Lower)).Dilate((int)(colorThresholds[3].h_Lower));
*
* Hsv c_Lower = new Hsv(colorThresholds[3].h_Lower, colorThresholds[3].s_Lower, colorThresholds[3].v_Lower);
* Hsv c_Upper = new Hsv(colorThresholds[3].h_Upper, colorThresholds[3].s_Upper, colorThresholds[3].v_Upper);
* Image<Gray, byte> findEdges = frame.ToImage<Gray, Byte>();
* findEdges = img.InRange(c_Lower, c_Upper).Erode(2);
*
* Gray cannyThreshold = new Gray(180);
* Gray cannyThresholdLinking = new Gray(120);
* Gray circleAccumulatorThreshold = new Gray(60);
*
* CircleF[] foundCircles = findEdges.HoughCircles(
* cannyThreshold,
* circleAccumulatorThreshold,
* 4.0, //Resolution of the accumulator used to detect centers of the circles
* 4.0, //Resolution of the accumulator used to detect centers of the circles
* 10.0, //min distance
* 10, //min radius
* 30 //max radius
* )[0]; //Get the circles from the first channel
*
* drawnImage = findEdges.Convert<Bgr, Byte>();
*
*
* // Find Largest Circle and Draw it on Frame
* for (int j = 0; j < foundCircles.Length; j++)
* {
* if (j == 0)
* {
* // Skip 0 because its the largest circle at this time
* }
* else
* {
* if (foundCircles[j].Area > foundCircles[largestCircle].Area) // Check to see if this is the Largest Circle
* {
* largestCircle = j;
* }
* }
* }
* if (foundCircles.Length != 0)
* drawnImage.Draw(foundCircles[largestCircle], new Bgr(0, 255, 0), 3, LineType.EightConnected, 0); // Circle Draw
*
*/
// Grid Detection
VectorOfVectorOfPoint Contours = new VectorOfVectorOfPoint();
Mat Draw = new Mat();
Hsv g_Lower = new Hsv(colorThresholds[5].h_Lower, colorThresholds[5].s_Lower, colorThresholds[5].v_Lower);
Hsv g_Upper = new Hsv(colorThresholds[5].h_Upper, colorThresholds[5].s_Upper, colorThresholds[5].v_Upper);
gridImg = img.InRange(g_Lower, g_Upper).Erode(2).Dilate(2);
CvInvoke.FindContours(coralNestImg, Contours, Draw, Emgu.CV.CvEnum.RetrType.Ccomp, ChainApproxMethod.ChainApproxSimple);
List <Shape> gShapes = new List <Shape>();
for (int j = 0; j < Contours.Size; j++)
{
double peramiter = CvInvoke.ArcLength(Contours[j], true);
VectorOfPoint positions = new VectorOfPoint();
CvInvoke.ApproxPolyDP(Contours[j], positions, 0.03 * peramiter, true);
var moments = CvInvoke.Moments((Contours[j]));
gShapes.Add(new Shape());
//Find Centroid of shape and area and store in shape class
gShapes[j].x_Center = (moments.M10 / moments.M00);
gShapes[j].y_Center = (moments.M01 / moments.M00);
gShapes[j].area = CvInvoke.ContourArea(Contours[j]);
}
if (gShapes.Count > 0)
{
for (int i = 0; i < gShapes.Count; i++)
{
if (gShapes[i].area > 400 || gShapes[i].area < 20000)
{
gShapes.RemoveAt(i); // remove unwanted shapes
}
}
/*
* // outline grid squares found
* gridDrawn = gridImg.Convert<Bgr, byte>();
* for(int i =0;i<gShapes.Count; i++)
* {
* Rectangle boundingRect = new Rectangle((int)(gShapes[i].x_Center - 50), (int)(gShapes[i].y_Center - 50), 100, 100);
* gridDrawn.Draw(boundingRect, new Bgr(0, 0, 255), 3, LineType.EightConnected, 0);
* }
*
* for(int i = 0; i < gShapes.Count; i++)
* {
* if (LastPosition.Count == 0)
* {
* LastPosition.Add(gShapes[i]); // store shapes positions for next rotation
* start = 0;
* end = foundShape.Count;
* //check if centroids are close to those of found shapes
* for (int j = 0; j < foundShape.Count; j++)
* {
* // Check centerpoints with tolerance of +- 20
* if(((foundShape[j].x_Center < gShapes[i].x_Center + 20) && (foundShape[j].x_Center > gShapes[i].x_Center - 20)) && ((foundShape[j].y_Center < gShapes[i].y_Center + 20) && (foundShape[j].y_Center > gShapes[i].y_Center - 20)))
* {
* key[i] = shapeType[j];
* }
* }
*
* }
* else
* {
* int tEnd = 0;
* for(int j = 0; j < gShapes.Count; j++)
* {
*
* if (((LastPosition[j].x_Center < gShapes[i].x_Center + 10) && (LastPosition[j].x_Center > gShapes[i].x_Center - 10)) && ((LastPosition[j].y_Center < gShapes[i].y_Center + 10) && (LastPosition[j].y_Center > gShapes[i].y_Center - 10)))
* {
* tEnd = j;
* }
* end = tEnd;
* start = (end - foundShape.Count);
*
* }
*
* for (int j = 0; j < gShapes.Count; j++)
* {
*
* if (((foundShape[j].x_Center < gShapes[i].x_Center + 20) && (foundShape[j].x_Center > gShapes[i].x_Center - 20)) && ((foundShape[j].y_Center < gShapes[i].y_Center + 20) && (foundShape[j].y_Center > gShapes[i].y_Center - 20)))
* {
* key[i + start] = shapeType[j];
* }
*
* }
* }
* }
*/
} // Grid analysis
// Display Images
pictureBox0.Image = frame.Bitmap;
pictureBox1.Image = coralNestDrawn.Bitmap;
pictureBox2.Image = coralOutcropDrawn.Bitmap;
pictureBox3.Image = starfishDrawn.Bitmap;
pictureBox4.Image = drawnImage.Bitmap;
pictureBox5.Image = laneImg.Bitmap;
GridBox.Image = gridDrawn.Bitmap;
if (!Sleep(20))
{
this.Invoke(new Action(() => { Close(); }));
break;
}
}
}
private void FindLicensePlate(
VectorOfVectorOfPoint contours,
int[,] hierachy,
int idx,
IInputArray gray,
IInputArray canny,
List <IInputOutputArray> licensePlateImagesList,
List <IInputOutputArray> filteredLicensePlateImagesList,
List <RotatedRect> detectedLicensePlateRegionList,
List <String> licenses)
{
if (hierachy.Length != 0)
{
for (; idx >= 0; idx = hierachy[idx, 0])
{
int numberOfChildren = GetNumberOfChildren(hierachy, idx);
//if it does not contains any children (charactor), it is not a license plate region
if (numberOfChildren == 0)
{
continue;
}
using (VectorOfPoint contour = contours[idx])
{
if (CvInvoke.ContourArea(contour) > 400)
{
if (numberOfChildren < 3)
{
//If the contour has less than 3 children, it is not a license plate (assuming license plate has at least 3 charactor)
//However we should search the children of this contour to see if any of them is a license plate
FindLicensePlate(contours, hierachy, hierachy[idx, 2], gray, canny, licensePlateImagesList,
filteredLicensePlateImagesList, detectedLicensePlateRegionList, licenses);
continue;
}
RotatedRect box = CvInvoke.MinAreaRect(contour);
if (box.Angle < -45.0)
{
float tmp = box.Size.Width;
box.Size.Width = box.Size.Height;
box.Size.Height = tmp;
box.Angle += 90.0f;
}
else if (box.Angle > 45.0)
{
float tmp = box.Size.Width;
box.Size.Width = box.Size.Height;
box.Size.Height = tmp;
box.Angle -= 90.0f;
}
double whRatio = (double)box.Size.Width / box.Size.Height;
if (!(3.0 < whRatio && whRatio < 10.0))
//if (!(1.0 < whRatio && whRatio < 2.0))
{
//if the width height ratio is not in the specific range,it is not a license plate
//However we should search the children of this contour to see if any of them is a license plate
//Contour<Point> child = contours.VNext;
if (hierachy[idx, 2] > 0)
{
FindLicensePlate(contours, hierachy, hierachy[idx, 2], gray, canny, licensePlateImagesList,
filteredLicensePlateImagesList, detectedLicensePlateRegionList, licenses);
}
continue;
}
using (UMat tmp1 = new UMat())
using (UMat tmp2 = new UMat())
{
PointF[] srcCorners = box.GetVertices();
PointF[] destCorners = new PointF[] {
new PointF(0, box.Size.Height - 1),
new PointF(0, 0),
new PointF(box.Size.Width - 1, 0),
new PointF(box.Size.Width - 1, box.Size.Height - 1)
};
using (Mat rot = CvInvoke.GetAffineTransform(srcCorners, destCorners))
{
CvInvoke.WarpAffine(gray, tmp1, rot, Size.Round(box.Size));
}
//resize the license plate such that the front is ~ 10-12. This size of front results in better accuracy from tesseract
Size approxSize = new Size(240, 180);
double scale = Math.Min(approxSize.Width / box.Size.Width, approxSize.Height / box.Size.Height);
Size newSize = new Size((int)Math.Round(box.Size.Width * scale), (int)Math.Round(box.Size.Height * scale));
CvInvoke.Resize(tmp1, tmp2, newSize, 0, 0, Inter.Cubic);
//removes some pixels from the edge
int edgePixelSize = 3;
Rectangle newRoi = new Rectangle(new Point(edgePixelSize, edgePixelSize),
tmp2.Size - new Size(2 * edgePixelSize, 2 * edgePixelSize));
UMat plate = new UMat(tmp2, newRoi);
UMat filteredPlate = FilterPlate(plate);
//Tesseract.Character[] words;
StringBuilder strBuilder = new StringBuilder();
using (UMat tmp = filteredPlate.Clone())
{
Emgu.CV.OCR.Tesseract.Character[] words;
_ocr.Recognize(tmp);
strBuilder.Append(_ocr.GetText());
words = _ocr.GetCharacters();
if (words.Length == 0)
{
continue;
}
for (int i = 0; i < words.Length; i++)
{
strBuilder.Append(words[i].Text);
}
}
licenses.Add(strBuilder.ToString());
//изображения номеров
licensePlateImagesList.Add(plate);
filteredLicensePlateImagesList.Add(filteredPlate);
detectedLicensePlateRegionList.Add(box);
}
}
}
}
}
}
public static void Run(Options options)
{
//load the image and compute the ratio of the old height
//to the new height, clone it, and resize it
using (var disposer = new Disposer())
{
var image = new Image <Bgr, byte>(options.Image);
disposer.Add(image);
Image <Bgr, byte> orig = image.Clone();
disposer.Add(orig);
double ratio = image.Height / 500.0;
image = ImageUtil.Resize(image, height: 500);
disposer.Add(image);
Image <Gray, byte> gray = image.Convert <Gray, byte>();
disposer.Add(gray);
gray = gray.SmoothGaussian(5);
disposer.Add(gray);
Image <Gray, byte> edged = gray.Canny(75, 200);
disposer.Add(edged);
Console.WriteLine("STEP 1: Edge Detection");
CvInvoke.Imshow("Image", image);
CvInvoke.Imshow("Edged", edged);
CvInvoke.WaitKey();
CvInvoke.DestroyAllWindows();
//find the contours in the edged image, keeping only the
//largest ones, and initialize the screen contour
VectorOfVectorOfPoint cnts = new VectorOfVectorOfPoint();
disposer.Add(cnts);
using (Image <Gray, byte> edgedClone = edged.Clone())
{
CvInvoke.FindContours(edgedClone, cnts, null, RetrType.List, ChainApproxMethod.ChainApproxSimple);
}
Point[] screenCnt = null;
foreach (VectorOfPoint c in
Enumerable.Range(0, cnts.Size).Select(i => cnts[i]).OrderByDescending(c => CvInvoke.ContourArea(c)).Take(5))
{
//approximate the contour
double peri = CvInvoke.ArcLength(c, true);
using (VectorOfPoint approx = new VectorOfPoint())
{
CvInvoke.ApproxPolyDP(c, approx, 0.02 * peri, true);
if (approx.Size == 4)
{
screenCnt = approx.ToArray();
break;
}
}
}
if (screenCnt == null)
{
Console.WriteLine("Failed to find polygon with four points");
return;
}
//show the contour (outline) of the piece of paper
Console.WriteLine("STEP 2: Find contours of paper");
image.Draw(screenCnt, new Bgr(0, 255, 0), 2);
CvInvoke.Imshow("Outline", image);
CvInvoke.WaitKey();
CvInvoke.DestroyAllWindows();
//apply the four point transform to obtain a top-down
//view of the original image
Image <Bgr, byte> warped = FourPointTransform(orig, screenCnt.Select(pt => new PointF((int)(pt.X * ratio), (int)(pt.Y * ratio))));
disposer.Add(warped);
//convert the warped image to grayscale, then threshold it
//to give it that 'black and white' paper effect
Image <Gray, byte> warpedGray = warped.Convert <Gray, byte>();
disposer.Add(warpedGray);
warpedGray = warpedGray.ThresholdAdaptive(new Gray(251), AdaptiveThresholdType.GaussianC, ThresholdType.Binary, 251, new Gray(10));
disposer.Add(warpedGray);
Console.WriteLine("STEP 3: Apply perspective transform");
Image <Bgr, byte> origResized = ImageUtil.Resize(orig, height: 650);
disposer.Add(origResized);
CvInvoke.Imshow("Original", origResized);
Image <Gray, byte> warpedResized = ImageUtil.Resize(warpedGray, height: 650);
disposer.Add(warpedResized);
CvInvoke.Imshow("Scanned", warpedResized);
CvInvoke.WaitKey();
CvInvoke.DestroyAllWindows();
}
}
// Update is called once per frame
void Update()
{
//CvInvoke.CvtColor(image, imgGray, ColorConversion.Bgr2Gray);
Mat image;
Mat structElt = CvInvoke.GetStructuringElement(eltShape, new Size(sizeX, sizeY), new Point(coordX, coordY));
//Query the frame
if (video.IsOpened)
{
image = video.QueryFrame();
//HSV img
Mat imgHSV = image.Clone();
CvInvoke.CvtColor(image, imgHSV, ColorConversion.Bgr2Hsv);
//Blur
Mat imgHSVBlur = imgHSV.Clone();
CvInvoke.MedianBlur(imgHSVBlur, imgHSVBlur, 21);
//New Img
Image <Hsv, Byte> newImg = imgHSVBlur.ToImage <Hsv, Byte>();
Hsv lowerBound = new Hsv(cMin, sMin, vMin);
Hsv higherBound = new Hsv(cMax, sMax, vMax);
Image <Gray, Byte> thresholdImg = newImg.InRange(lowerBound, higherBound);
Image <Gray, Byte> thresholdImgErode = thresholdImg.Clone();
CvInvoke.Dilate(thresholdImgErode, thresholdImgErode, structElt, new Point(-1, -1), nbrIteration, border, new MCvScalar(0));
CvInvoke.Erode(thresholdImgErode, thresholdImgErode, structElt, new Point(-1, -1), nbrIteration, border, new MCvScalar(0));
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint();
VectorOfPoint biggestContour = new VectorOfPoint();
int biggestContourIndex = 0;
double biggestContourArea = 0;
Mat hierarchy = new Mat();
CvInvoke.FindContours(thresholdImgErode, contours, hierarchy, retraitType, chainApprox);
for (int i = 0; i < contours.Size; i++)
{
if (biggestContourArea < CvInvoke.ContourArea(contours[i]))
{
biggestContourArea = CvInvoke.ContourArea(contours[i]);
biggestContour = contours[i];
biggestContourIndex = i;
}
}
/*
* if (biggestContour.Size > 0)
* {
* int x = 0;
* int y = 0;
*
* for (int i = 0; i < biggestContour.Size; i++)
* {
* x += biggestContour[i].X;
* y += biggestContour[i].Y;
* }
*
* x /= biggestContour.Size;
* y /= biggestContour.Size;
*
* Point centroid = new Point(x, y);
*
* CvInvoke.Circle(image, centroid, 10, new MCvScalar(0, 0, 255));
* }*/
//Centroid
var moments = CvInvoke.Moments(biggestContour);
int cx = (int)(moments.M10 / moments.M00);
int cy = (int)(moments.M01 / moments.M00);
Point centroid = new Point(cx, cy);
CvInvoke.Circle(image, centroid, 10, new MCvScalar(0, 0, 255));
CvInvoke.DrawContours(image, contours, biggestContourIndex, new MCvScalar(0, 0, 0));
//Invoke C++ interface fonction "Imshow"
CvInvoke.Imshow("Video view", image);
//Invoke C++ interface fonction "Imshow"
CvInvoke.Imshow("Video view hsv", imgHSVBlur);
// Invoke C++ interface fonction "Imshow"
CvInvoke.Imshow("Video view threshhold", thresholdImg);
//Invoke C++ interface fonction "Imshow"
CvInvoke.Imshow("Video view threshold erode + dilate", thresholdImgErode);
//Block thread for 24 milisecond
CvInvoke.WaitKey(24);
}
}
private void findShapes(Image <Bgr, byte> imgInput)
{
Image <Gray, byte> imgCanny = new Image <Gray, byte>(imgInput.Width, imgInput.Height, new Gray(0));
imgCanny = imgInput.Canny(20, 50);
//pictureBox1.Image = imgCanny.Bitmap;
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint();
Mat mat = new Mat();
CvInvoke.FindContours(imgCanny, contours, mat, Emgu.CV.CvEnum.RetrType.Tree, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxSimple);
int[] shapes = new int[4];
/* [0] . . . Triangle
* [1] . . . Square
* [2] . . . Rectangle
* [3] . . . Circle
*/
for (int i = 0; i < contours.Size; i++)
{
double perimeter = CvInvoke.ArcLength(contours[i], true);
VectorOfPoint approx = new VectorOfPoint();
CvInvoke.ApproxPolyDP(contours[i], approx, 0.04 * perimeter, true);
CvInvoke.DrawContours(imgInput, contours, i, new MCvScalar(0, 0, 255), 2);
pictureBox1.Image = imgInput.Bitmap;
//moments center of the shape
var moments = CvInvoke.Moments(contours[i]);
int x = (int)(moments.M10 / moments.M00);
int y = (int)(moments.M01 / moments.M00);
if (CvInvoke.ContourArea(approx) > 250) //ONLY CONSIDERS SHAPES WITH AREA GREATER THAN 100
{
if (approx.Size == 3)
{
shapes[0]++;
CvInvoke.PutText(imgInput, "Triangle", new Point(x, y),
Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 0), 1);
}
if (approx.Size == 4)
{
Rectangle rect = CvInvoke.BoundingRectangle(contours[i]);
double ar = (double)rect.Width / rect.Height;
if (ar >= 0.95 && ar <= 1.05)
{
shapes[1]++;
CvInvoke.PutText(imgInput, "Square", new Point(x, y),
Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 0), 1);
}
else
{
shapes[2]++;
CvInvoke.PutText(imgInput, "Rectangle", new Point(x, y),
Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 0), 1);
}
}
if (approx.Size > 6)
{
CvInvoke.PutText(imgInput, "Circle", new Point(x, y),
Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 0), 1);
shapes[3]++;
}
pictureBox1.Image = imgInput.Bitmap;
}
}
textBox8.Text = shapes[0].ToString();
textBox6.Text = shapes[1].ToString();
textBox4.Text = shapes[2].ToString();
textBox2.Text = shapes[3].ToString();
}
private Mat DetectObject(Mat detectionFrame, Mat displayFrame, Rectangle box)
{
Image <Bgr, Byte> buffer_im = displayFrame.ToImage <Bgr, Byte>();
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
IOutputArray hirarchy = null;
// Construir lista de contur (contornos)
CvInvoke.FindContours(detectionFrame, contours, hirarchy, RetrType.List, ChainApproxMethod.ChainApproxSimple);
// seleccionar el contour (contorno) mas grande
if (contours.Size > 0)
{
double maxArea = 0;
int chosen = 0;
VectorOfPoint contour = null;
for (int i = 0; i < contours.Size; i++)
{
contour = contours[i];
double area = CvInvoke.ContourArea(contour);
if (area > maxArea)
{
maxArea = area;
chosen = i;
}
}
VectorOfPoint hullPoints = new VectorOfPoint();
VectorOfInt hullInt = new VectorOfInt();
CvInvoke.ConvexHull(contours[chosen], hullPoints, true);
CvInvoke.ConvexHull(contours[chosen], hullInt, false);
Mat defects = new Mat();
if (hullInt.Size > 3)
{
CvInvoke.ConvexityDefects(contours[chosen], hullInt, defects);
}
box = CvInvoke.BoundingRectangle(hullPoints);
CvInvoke.Rectangle(displayFrame, box, drawingColor); //Box rectangulo que encierra el area mas grande
center = new Point(box.X + box.Width / 2, box.Y + box.Height / 2); //centro rectangulo MOUSE
var infoCentro = new string[] { $"Centro", $"Posicion: {center.X}, {center.Y}" };
WriteMultilineText(displayFrame, infoCentro, new Point(center.X + 30, center.Y));
CvInvoke.Circle(displayFrame, new Point(center.X, center.Y), 2, new MCvScalar(0, 100, 0), 4);
detectGesture = true;
buffer_im.Dispose();
defects.Dispose();
detectionFrame.Dispose();
return(displayFrame);
}
buffer_im.Dispose();
detectionFrame.Dispose();
detectGesture = false;
return(displayFrame);
}
}
private void button2_Click(object sender, EventArgs e)
{
Image triangleRectangleImageBox1;
Image circleImageBox1;
Image lineImageBox1;
StringBuilder msgBuilder = new StringBuilder("Performance: ");
Bitmap clip_bmp = new Bitmap(cliped_image);
//Load the image from file and resize it for display
Image <Bgr, Byte> img = new Image <Bgr, byte>(clip_bmp).Resize(400, 400, Emgu.CV.CvEnum.Inter.Linear, true);
//Convert the image to grayscale and filter out the noise
UMat uimage = new UMat();
CvInvoke.CvtColor(img, uimage, ColorConversion.Bgr2Gray);
//use image pyr to remove noise
UMat pyrDown = new UMat();
CvInvoke.PyrDown(uimage, pyrDown);
CvInvoke.PyrUp(pyrDown, uimage);
#region circle detection
Stopwatch watch = Stopwatch.StartNew();
double cannyThreshold = 180.0; // 180
double circleAccumulatorThreshold = 120.0; // 120
CircleF[] circles = CvInvoke.HoughCircles(uimage, HoughType.Gradient, 2.0, 20.0, cannyThreshold, circleAccumulatorThreshold, 5);
watch.Stop();
msgBuilder.Append(String.Format("Hough circles - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
#region Canny and edge detection
watch.Reset(); watch.Start();
double cannyThresholdLinking = 120.0; // 120
UMat cannyEdges = new UMat();
CvInvoke.Canny(uimage, cannyEdges, cannyThreshold, cannyThresholdLinking);
LineSegment2D[] lines = CvInvoke.HoughLinesP(
cannyEdges,
1, //Distance resolution in pixel-related units - 1
Math.PI / 45.0, //Angle resolution measured in radians.
0, //threshold - 20
0, //min Line width - 30
10); //gap between lines - 10
watch.Stop();
msgBuilder.Append(String.Format("Canny & Hough lines - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
#region Find triangles and rectangles
watch.Reset(); watch.Start();
List <Triangle2DF> triangleList = new List <Triangle2DF>();
List <RotatedRect> boxList = new List <RotatedRect>(); //a box is a rotated rectangle
using (Emgu.CV.Util.VectorOfVectorOfPoint contours = new Emgu.CV.Util.VectorOfVectorOfPoint())
{
CvInvoke.FindContours(cannyEdges, contours, null, RetrType.List, ChainApproxMethod.ChainApproxSimple);
int count = contours.Size;
for (int i = 0; i < count; i++)
{
using (Emgu.CV.Util.VectorOfPoint contour = contours[i])
using (Emgu.CV.Util.VectorOfPoint approxContour = new Emgu.CV.Util.VectorOfPoint())
{
CvInvoke.ApproxPolyDP(contour, approxContour, CvInvoke.ArcLength(contour, true) * 0.05, true);
if (CvInvoke.ContourArea(approxContour, false) > 250) //only consider contours with area greater than 250
{
if (approxContour.Size == 3) //The contour has 3 vertices, it is a triangle
{
Point[] pts = approxContour.ToArray();
triangleList.Add(new Triangle2DF(
pts[0],
pts[1],
pts[2]
));
}
else if (approxContour.Size == 4) //The contour has 4 vertices.
{
#region determine if all the angles in the contour are within [80, 100] degree
bool isRectangle = true;
Point[] pts = approxContour.ToArray();
LineSegment2D[] edges = PointCollection.PolyLine(pts, true);
for (int j = 0; j < edges.Length; j++)
{
double angle = Math.Abs(
edges[(j + 1) % edges.Length].GetExteriorAngleDegree(edges[j]));
if (angle < 80 || angle > 100)
{
isRectangle = false;
break;
}
}
#endregion
if (isRectangle)
{
boxList.Add(CvInvoke.MinAreaRect(approxContour));
}
}
}
}
}
}
watch.Stop();
msgBuilder.Append(String.Format("Triangles & Rectangles - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
//originalImageBox.Image = img.ToBitmap();
this.Text = msgBuilder.ToString();
#region draw triangles and rectangles
Image <Bgr, Byte> triangleRectangleImage = img.CopyBlank();
foreach (Triangle2DF triangle in triangleList)
{
triangleRectangleImage.Draw(triangle, new Bgr(Color.DarkBlue), 2);
}
foreach (RotatedRect box in boxList)
{
triangleRectangleImage.Draw(box, new Bgr(Color.DarkOrange), 2);
}
triangleRectangleImageBox1 = triangleRectangleImage.ToBitmap();
#endregion
#region draw circles
Image <Bgr, Byte> circleImage = img.CopyBlank();
foreach (CircleF circle in circles)
{
circleImage.Draw(circle, new Bgr(Color.Brown), 2);
}
circleImageBox1 = circleImage.ToBitmap();
#endregion
#region draw lines
Image <Bgr, Byte> lineImage = img.CopyBlank();
foreach (LineSegment2D line in lines)
{
lineImage.Draw(line, new Bgr(Color.Green), 2);
}
lineImageBox1 = lineImage.ToBitmap();
#endregion
cv_image = lineImageBox1;
pictureBox2.Image = lineImageBox1;
}
/// <summary>
/// Funkcja wykrywająca znaki
/// </summary>
private void WykryjZnaki()
{
// lista trójkątów
List <Triangle2DF> triangleList = new List <Triangle2DF>();
// lista prostokątów i kwadratów
List <RotatedRect> boxList = new List <RotatedRect>();
ZwiekszProgressBar(1);
// przetworzenie zdjecia do postaci wskazującej tylko białe kontury na czarnym tle
Image <Gray, byte> canny_zdj = new Image <Gray, byte>(imgInput.Width, imgInput.Height, new Gray(0));
canny_zdj = imgInput.Canny(300, 250);
// przypisanie canny_zdj do pictureBox i rozciagniecie
zdjecieCannyBox.Image = canny_zdj.Bitmap;
zdjecieCannyBox.SizeMode = PictureBoxSizeMode.StretchImage;
ZwiekszProgressBar(2);
LineSegment2D[] lines = CvInvoke.HoughLinesP(
canny_zdj,
1,
Math.PI / 45.0,
20,
30,
10);
Image <Gray, byte> imgOut = canny_zdj.Convert <Gray, byte>().ThresholdBinary(new Gray(50), new Gray(200));
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint();
Mat hier = new Mat();
ZwiekszProgressBar(1);
// wygładzenie obrazu
imgSmooth = imgInput.PyrDown().PyrUp();
imgSmooth._SmoothGaussian(3);
// ograniczenie wykrywanych figur do odpowiedniego zakresu ze skali RGB - zółtego
imgOut = imgSmooth.InRange(new Bgr(0, 140, 150), new Bgr(80, 255, 255));
imgOut = imgOut.PyrDown().PyrUp();
imgOut._SmoothGaussian(3);
ZwiekszProgressBar(2);
Dictionary <int, double> dict = new Dictionary <int, double>();
// wyszukanie konturów spełniajacych wymogi odnośnie między innymi koloru
CvInvoke.FindContours(imgOut, contours, null, Emgu.CV.CvEnum.RetrType.List, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxSimple);
label1.Text = contours.Size.ToString();
// jeśli odnaleziono choćby jeden kontur
if (contours.Size > 0)
{
// petla przechodząca po wszystkich wykrytych konturach
for (int i = 0; i < contours.Size; i++)
{
using (VectorOfPoint contour = contours[i])
using (VectorOfPoint approxContour = new VectorOfPoint())
{
CvInvoke.ApproxPolyDP(contour, approxContour, CvInvoke.ArcLength(contour, true) * 0.05, true);
// filtr wielkości pola wykrytego konturu
if (CvInvoke.ContourArea(approxContour, false) > 50)
{
// jesli to trójkąt
if (approxContour.Size == 3)
{
// tablica punktów i dodanie ich do tablicy trójkątów
Point[] pts = approxContour.ToArray();
triangleList.Add(new Triangle2DF(
pts[0],
pts[1],
pts[2]
));
// sprawdzenie, czy wykryty trojkat jest figurą obróconą jednym z wierzcholkow do dołu
if (pts[1].X > pts[0].X && pts[1].Y > pts[0].Y)
{
// ustawienie znaku A-7
UstawWykrytyZnak(2);
double area = CvInvoke.ContourArea(contours[i]);
// dodanie do tablicy (słownika) głównej
dict.Add(i, area);
}
}
// jesli to czworokąt
else if (approxContour.Size == 4)
{
bool isRectangle = true;
// rozbicie figury na pojedyncze krawędzie
Point[] pts = approxContour.ToArray();
LineSegment2D[] edges = PointCollection.PolyLine(pts, true);
// petla przechodzaca po wszystkich krawedziach
for (int j = 0; j < edges.Length; j++)
{
// sprawdzenie wielkosci kąta miedzy sprawdzanymi krawędziami
double angle = Math.Abs(
edges[(j + 1) % edges.Length].GetExteriorAngleDegree(edges[j]));
// przerwanie jeśli kąty w figurze są mniejsze niż 80 i wieksze niż 100 stopni
if (angle < 80 || angle > 100)
{
isRectangle = false;
break;
}
}
if (isRectangle)
{
RotatedRect rrect = CvInvoke.MinAreaRect(contours[i]);
// ostateczne sprawdzenie, czy wykryta figura jest obrócona względem środka o wartość od 40 do 50 stopni
// znak D-1 jest obróconym kwadratem o 45 stopni względem środka
if ((rrect.Angle < -40 && rrect.Angle > -50) || (rrect.Angle > 40 && rrect.Angle < 50))
{
boxList.Add(CvInvoke.MinAreaRect(approxContour));
double area = CvInvoke.ContourArea(contours[i]);
dict.Add(i, area);
UstawWykrytyZnak(1);
}
}
}
}
}
}
}
ZwiekszProgressBar(2);
var item = dict.OrderByDescending(v => v.Value);
foreach (var it in item)
{
int key = int.Parse(it.Key.ToString());
// pobranie odpowiednich konturów
Rectangle rect = CvInvoke.BoundingRectangle(contours[key]);
// narysowanie czerwonego prostokąta wokół wykrytego znaku
CvInvoke.Rectangle(imgInput, rect, new MCvScalar(0, 0, 255), 1);
}
ZwiekszProgressBar(2);
pictureBox2.Image = imgInput.Bitmap;
pictureBox2.SizeMode = PictureBoxSizeMode.StretchImage;
// utworzenie zdjęcia wskazującego WSZYSTKIE kontury w początkowym zdjęciu - czerowne linie
Image <Bgr, Byte> lineImage = imgInput.CopyBlank();
foreach (LineSegment2D line in lines)
{
lineImage.Draw(line, new Bgr(Color.Red), 1);
}
zdjecieWykrytyZnak.Image = lineImage.Bitmap;
zdjecieWykrytyZnak.SizeMode = PictureBoxSizeMode.StretchImage;
}
//
//COMIENZAN FUNCIONES DE EDDIE
//
private void DetectObject(Mat detectionFrame, Mat displayFrame)
{
System.Drawing.Rectangle box = new System.Drawing.Rectangle();
Image <Bgr, byte> temp = detectionFrame.ToImage <Bgr, byte>();
temp = temp.Rotate(180, new Bgr(0, 0, 0));
Image <Bgr, Byte> buffer_im = displayFrame.ToImage <Bgr, Byte>();
float a = buffer_im.Width;
float b = buffer_im.Height;
MessageBox.Show("El tamano camara es W: " + a.ToString() + " y H:" + b.ToString());
boxList.Clear();
rect.Clear();
triangleList.Clear();
circleList.Clear();
ellipseList.Clear();
//transforma imagen
//UMat uimage = new UMat();
// CvInvoke.CvtColor(displayFrame, uimage, ColorConversion.Bgr2Gray);
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
/// IOutputArray hirarchy = null;
/// CvInvoke.FindContours(detectionFrame, contours, hirarchy, RetrType.External, ChainApproxMethod.ChainApproxSimple);
///CvInvoke.Polylines(detectionFrame, contours, true, new MCvScalar(255, 0, 0), 2, LineType.FourConnected);
Image <Bgr, Byte> resultadoFinal = displayFrame.ToImage <Bgr, byte>();
resultadoFinal = resultadoFinal.Rotate(180, new Bgr(0, 0, 0));
//Circulos
//double cannyThreshold = 180.0;
//double circleAccumulatorThreshold = 120;
//CircleF[] circles = CvInvoke.HoughCircles(detectionFrame, HoughType.Gradient, 2.0, 20.0, cannyThreshold, circleAccumulatorThreshold, 5);
/// if (contours.Size > 0)
///{
double maxArea = 1000;
int chosen = 0;
VectorOfPoint contour = null;
/*
* for (int i = 0; i < contours.Size; i++)
* {
* contour = contours[i];
*
* double area = CvInvoke.ContourArea(contour);
* if (area > maxArea)
* {
* System.Drawing.Rectangle rect = new System.Drawing.Rectangle();
* // maxArea = area;
* chosen = i;
* //}
* //}
*
* //Boxes
* VectorOfPoint hullPoints = new VectorOfPoint();
* VectorOfInt hullInt = new VectorOfInt();
*
* CvInvoke.ConvexHull(contours[chosen], hullPoints, true);
* CvInvoke.ConvexHull(contours[chosen], hullInt, false);
*
* Mat defects = new Mat();
*
* if (hullInt.Size > 3)
* CvInvoke.ConvexityDefects(contours[chosen], hullInt, defects);
*
* box = CvInvoke.BoundingRectangle(hullPoints);
* CvInvoke.Rectangle(displayFrame, box, drawingColor);//Box rectangulo que encierra el area mas grande
* // cropbox = crop_color_frame(displayFrame, box);
*
* buffer_im.ROI = box;
*
* Image<Bgr, Byte> cropped_im = buffer_im.Copy();
* //pictureBox8.Image = cropped_im.Bitmap;
* System.Drawing.Point center = new System.Drawing.Point(box.X + box.Width / 2, box.Y + box.Height / 2);//centro rectangulo MOUSE
* System.Drawing.Point esquina_superiorI = new System.Drawing.Point(box.X, box.Y);
* System.Drawing.Point esquina_superiorD = new System.Drawing.Point(box.Right, box.Y);
* System.Drawing.Point esquina_inferiorI = new System.Drawing.Point(box.X, box.Y + box.Height);
* System.Drawing.Point esquina_inferiorD = new System.Drawing.Point(box.Right, box.Y + box.Height);
* CvInvoke.Circle(displayFrame, esquina_superiorI, 5, new MCvScalar(0, 0, 255), 2);
* CvInvoke.Circle(displayFrame, esquina_superiorD, 5, new MCvScalar(0, 0, 255), 2);
* CvInvoke.Circle(displayFrame, esquina_inferiorI, 5, new MCvScalar(0, 0, 255), 2);
* CvInvoke.Circle(displayFrame, esquina_inferiorD, 5, new MCvScalar(0, 0, 255), 2);
* CvInvoke.Circle(displayFrame, center, 5, new MCvScalar(0, 0, 255), 2);
* VectorOfPoint start_points = new VectorOfPoint();
* VectorOfPoint far_points = new VectorOfPoint();
*
*
*
*
*
*
* }
* }
*/
//Dibuja borde rojo
var temp2 = temp.SmoothGaussian(5).Convert <Gray, byte>().ThresholdBinary(new Gray(20), new Gray(255));
temp2 = temp2.Rotate(180, new Gray(0));
VectorOfVectorOfPoint contorno = new VectorOfVectorOfPoint();
Mat mat = new Mat();
CvInvoke.FindContours(temp2, contorno, mat, Emgu.CV.CvEnum.RetrType.External, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxSimple);
for (int i = 0; i < contorno.Size; i++)
{
VectorOfPoint approxContour = new VectorOfPoint();
double perimetro = CvInvoke.ArcLength(contorno[i], true);
VectorOfPoint approx = new VectorOfPoint();
VectorOfPointF approxF = new VectorOfPointF();
double area = CvInvoke.ContourArea(contorno[i]);
if (area > 5000)
{
CvInvoke.ApproxPolyDP(contorno[i], approx, 0.04 * perimetro, true);
// CvInvoke.DrawContours(displayFrame, contorno, i, new MCvScalar(255, 0, 0), 2);
//pictureBox4.Image = temp2.Bitmap;
var moments = CvInvoke.Moments(contorno[i]);
int x = (int)(moments.M10 / moments.M00);
int y = (int)(moments.M01 / moments.M00);
resultados.Add(approx);
bool isShape;
if (approx.Size == 3) //The contour has 3 vertices, it is a triangle
{
System.Drawing.Point[] pts = approx.ToArray();
double perimetro2 = CvInvoke.ArcLength(contorno[i], true);
double area2 = CvInvoke.ContourArea(contorno[i]);
double circularidad = 4 * Math.PI * area2 / Math.Pow(perimetro2, 2);
MessageBox.Show("circularidad triangulo" + circularidad);
MessageBox.Show("Es triangulo ");
/*Triangle2DF triangle = new Triangle2DF(pts[0], pts[1], pts[2]);
* resultadoFinal.Draw(triangle, new Bgr(System.Drawing.Color.Cyan), 1);
* CvInvoke.DrawContours(resultadoFinal, contorno, i, new MCvScalar(255, 255, 255), 1, LineType.AntiAlias);
* CvInvoke.PutText(resultadoFinal, "Triangle", new System.Drawing.Point(x, y),
* Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 255), 2);
* resTri.Add(approx);*/
//MessageBox.Show("No es triangulo ");
//Triangle2DF triangle = new Triangle2DF(pts[0], pts[1], pts[2]);
//resultadoFinal.Draw(triangle, new Bgr(System.Drawing.Color.Red), 2);
RotatedRect rectangle = CvInvoke.MinAreaRect(approx);
CvInvoke.DrawContours(resultadoFinal, contorno, i, new MCvScalar(255, 255, 255), 1, LineType.AntiAlias);
resultadoFinal.Draw(rectangle, new Bgr(System.Drawing.Color.Cyan), 1);
rect.Add(CvInvoke.BoundingRectangle(approx));
}
if (approx.Size == 4) //The contour has 4 vertices.
{
//RotatedRect tt = new RotatedRect(CvInvoke.MinAreaRect(approx).Center, CvInvoke.MinAreaRect(approx).Size, 270) ;
//boxList.Add(tt);
//Calcular si es cuadrado
System.Drawing.Rectangle rectAux = CvInvoke.BoundingRectangle(contorno[i]);
double ar = (double)rectAux.Width / rectAux.Height;
//Calcular circularidad
double perimetro2 = CvInvoke.ArcLength(contorno[i], true);
double area2 = CvInvoke.ContourArea(contorno[i]);
double circularidad = 4 * Math.PI * area2 / Math.Pow(perimetro2, 2);
MessageBox.Show("circularidad rect " + circularidad);
if (circularidad > 0.69)
{
//Si la circularidad>0.6 y cumple proporcion es cuadrado
if (ar >= 0.8 && ar <= 1.0)
{
MessageBox.Show("Cuadrado ");
RotatedRect rectangle = CvInvoke.MinAreaRect(contorno[i]);
CvInvoke.DrawContours(resultadoFinal, contorno, i, new MCvScalar(255, 255, 255), 1, LineType.AntiAlias);
resultadoFinal.Draw(rectangle, new Bgr(System.Drawing.Color.Cyan), 1);
//CvInvoke.PutText(resultadoFinal, "Rectangle", new System.Drawing.Point(x, y),
//Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 255), 2);
rect.Add(CvInvoke.BoundingRectangle(approx));
}
//Es elipse
else
{
MessageBox.Show("parecia rectangulo pero era elipse ");
Ellipse final_ellipse = new Ellipse(CvInvoke.MinAreaRect(contorno[i]).Center, CvInvoke.MinAreaRect(contorno[i]).Size, 0);
Ellipse final_ellipseDibujo = new Ellipse(CvInvoke.MinAreaRect(contorno[i]).Center, CvInvoke.MinAreaRect(contorno[i]).Size, 90);
ellipseList.Add(final_ellipse);
//IConvexPolygonF poligono = CvInvoke.MinAreaRect(approx);
//resultadoFinal.Draw(poligono, new Bgr(Color.Cyan), 1);
resultadoFinal.Draw(final_ellipseDibujo, new Bgr(System.Drawing.Color.Cyan), 1);
CvInvoke.DrawContours(resultadoFinal, contorno, i, new MCvScalar(255, 255, 255), 1, LineType.AntiAlias);
//CvInvoke.PutText(resultadoFinal, "Figura circular", new System.Drawing.Point(x, y),
// Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 255), 2);
}
}
//Es rectangulo
else
{
RotatedRect rectangle = CvInvoke.MinAreaRect(contorno[i]);
CvInvoke.DrawContours(resultadoFinal, contorno, i, new MCvScalar(255, 255, 255), 1, LineType.AntiAlias);
resultadoFinal.Draw(rectangle, new Bgr(System.Drawing.Color.Cyan), 1);
//CvInvoke.PutText(resultadoFinal, "Rectangle", new System.Drawing.Point(x, y),
//Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 255), 2);
rect.Add(CvInvoke.BoundingRectangle(approx));
}
/* //prueba imagen de rectangulo
* //--------------------------------------PART 1 : DRAWING STUFF IN A BITMAP------------------------------------------------------------------------------------
* System.Drawing.Point[] pts = approx.ToArray();
*
* System.Drawing.PointF[] mypoints = Array.ConvertAll(
* pts.ToArray<System.Drawing.Point>(),
* value => new System.Drawing.PointF(value.X, value.Y)
* );
*
* System.Drawing.Rectangle r = new System.Drawing.Rectangle(0, 0, CvInvoke.BoundingRectangle(approx).Width, CvInvoke.BoundingRectangle(approx).Height);
* Pen blackPen = new Pen(System.Drawing.Color.FromArgb(255, 255, 0, 0), 1);
* bmp = new Bitmap(r.Width+100,r.Height+10, PixelFormat.Format32bppArgb);
* Graphics g = Graphics.FromImage(bmp);
* g.DrawRectangle(blackPen, r); //rectangle 1
* g.DrawPolygon(blackPen,mypoints);
* System.Drawing.Rectangle rcrop = new System.Drawing.Rectangle(r.X, r.Y, r.Width + 10, r.Height + 10);//This is the cropping rectangle (bonding box adding 10 extra units width and height)
*
* //Crop the model from the bmp
* Bitmap src = bmp;
* // Bitmap target = new Bitmap(r.Width, r.Height);
* //using (Graphics gs = Graphics.FromImage(target))
* //{
* // gs.DrawImage(src, rcrop, r, GraphicsUnit.Pixel);
* // gs.Dispose();
* //}
* //--------------------------------------PART 2 : SAVING THE BMP AS JPG------------------------------------------------------------------------------------
* src.Save("testOJO.jpg");*/
}
/* ELIMINAR
* if (approx.Size == 5 )
* {
* System.Drawing.Point[] pts = approx.ToArray();
*
* //MessageBox.Show("Cantidad puntos poligono "+pts.Length);
* //IConvexPolygonF poligono = CvInvoke.MinAreaRect(approx);
* //resultadoFinal.Draw(poligono, new Bgr(Color.Cyan), 1);
* CvInvoke.DrawContours(resultadoFinal, contorno, i, new MCvScalar(255, 255, 0), 1, LineType.AntiAlias);
* CvInvoke.PutText(resultadoFinal, "Pentagon", new System.Drawing.Point(x, y),
* Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 255), 2);
* }*/
if (approx.Size >= 5)
{
double perimetro2 = CvInvoke.ArcLength(contorno[i], true);
double area2 = CvInvoke.ContourArea(contorno[i]);
double circularidad = 4 * Math.PI * area2 / Math.Pow(perimetro2, 2);
MessageBox.Show("circularidad elipse " + circularidad);
Ellipse final_ellipse = new Ellipse(CvInvoke.MinAreaRect(contorno[i]).Center, CvInvoke.MinAreaRect(contorno[i]).Size, 0);
Ellipse final_ellipseDibujo = new Ellipse(CvInvoke.MinAreaRect(contorno[i]).Center, CvInvoke.MinAreaRect(contorno[i]).Size, 90);
ellipseList.Add(final_ellipse);
//IConvexPolygonF poligono = CvInvoke.MinAreaRect(approx);
//resultadoFinal.Draw(poligono, new Bgr(Color.Cyan), 1);
resultadoFinal.Draw(final_ellipseDibujo, new Bgr(System.Drawing.Color.Cyan), 1);
CvInvoke.DrawContours(resultadoFinal, contorno, i, new MCvScalar(255, 255, 255), 1, LineType.AntiAlias);
//CvInvoke.PutText(resultadoFinal, "Figura circular", new System.Drawing.Point(x, y),
// Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 255), 2);
}
/* _Eliminar
* if (approx.Size > 6)
* {
*
* double circularidad = 4 * Math.PI * area / (Math.Pow(2, perimetro));
* MessageBox.Show("circularidad circulo "+circularidad);
* CvInvoke.PutText(resultadoFinal, "Circle", new System.Drawing.Point(x, y),
* Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.5, new MCvScalar(0, 255, 255), 2);
* CircleF circle = CvInvoke.MinEnclosingCircle(approx);
* circleList.Add(circle);
* CvInvoke.DrawContours(resultadoFinal, contorno, i, new MCvScalar(255, 255, 255), 1, LineType.AntiAlias);
* resultadoFinal.Draw(circle, new Bgr(System.Drawing.Color.Cyan), 1);
*
*
*
* }*/
}
}
pictureBox2.Image = resultadoFinal.Bitmap;
button2.Enabled = true;
///}
}
}
/// <summary>
/// Get the piece corners by finding the biggest rectangle of the contour points
/// </summary>
/// <param name="pieceID">ID of the piece</param>
/// <param name="pieceImgBw">Black white image of piece</param>
/// <param name="pieceImgColor">Color image of piece</param>
/// <returns>List with corner points</returns>
public override List <Point> FindCorners(string pieceID, Bitmap pieceImgBw, Bitmap pieceImgColor)
{
PluginFactory.LogHandle.Report(new LogEventInfo(pieceID + " Finding corners by finding the maximum rectangle within candidate points"));
List <Point> corners = new List <Point>();
// Find all dominant corner points using the GFTTDetector (this uses the Harris corner detector)
GFTTDetector detector = new GFTTDetector(500, 0.01, 5, 2, true, 0.04);
MKeyPoint[] keyPoints = detector.Detect(new Image <Gray, byte>(pieceImgBw));
List <Point> possibleCorners = keyPoints.Select(k => Point.Round(k.Point)).ToList();
if (possibleCorners.Count > 0)
{
// Sort the dominant corners by the distance to upper left corner of the bounding rectangle (0, 0) and keep only the corners that are near enough to this point
List <Point> possibleCornersSortedUpperLeft = new List <Point>(possibleCorners);
possibleCornersSortedUpperLeft.Sort(new DistanceToPointComparer(new Point(0, 0), DistanceOrders.NEAREST_FIRST));
double minCornerDistUpperLeft = Utils.Distance(possibleCornersSortedUpperLeft[0], new PointF(0, 0));
possibleCornersSortedUpperLeft = possibleCornersSortedUpperLeft.Where(c => Utils.Distance(c, new PointF(0, 0)) < minCornerDistUpperLeft * PieceFindCornersMaxCornerDistRatio).ToList();
// Sort the dominant corners by the distance to upper right corner of the bounding rectangle (ImageWidth, 0) and keep only the corners that are near enough to this point
List <Point> possibleCornersSortedUpperRight = new List <Point>(possibleCorners);
possibleCornersSortedUpperRight.Sort(new DistanceToPointComparer(new Point(pieceImgBw.Width, 0), DistanceOrders.NEAREST_FIRST));
double minCornerDistUpperRight = Utils.Distance(possibleCornersSortedUpperRight[0], new PointF(pieceImgBw.Width, 0));
possibleCornersSortedUpperRight = possibleCornersSortedUpperRight.Where(c => Utils.Distance(c, new PointF(pieceImgBw.Width, 0)) < minCornerDistUpperRight * PieceFindCornersMaxCornerDistRatio).ToList();
// Sort the dominant corners by the distance to lower right corner of the bounding rectangle (ImageWidth, ImageHeight) and keep only the corners that are near enough to this point
List <Point> possibleCornersSortedLowerRight = new List <Point>(possibleCorners);
possibleCornersSortedLowerRight.Sort(new DistanceToPointComparer(new Point(pieceImgBw.Width, pieceImgBw.Height), DistanceOrders.NEAREST_FIRST));
double minCornerDistLowerRight = Utils.Distance(possibleCornersSortedLowerRight[0], new PointF(pieceImgBw.Width, pieceImgBw.Height));
possibleCornersSortedLowerRight = possibleCornersSortedLowerRight.Where(c => Utils.Distance(c, new PointF(pieceImgBw.Width, pieceImgBw.Height)) < minCornerDistLowerRight * PieceFindCornersMaxCornerDistRatio).ToList();
// Sort the dominant corners by the distance to lower left corner of the bounding rectangle (0, ImageHeight) and keep only the corners that are near enough to this point
List <Point> possibleCornersSortedLowerLeft = new List <Point>(possibleCorners);
possibleCornersSortedLowerLeft.Sort(new DistanceToPointComparer(new Point(0, pieceImgBw.Height), DistanceOrders.NEAREST_FIRST));
double minCornerDistLowerLeft = Utils.Distance(possibleCornersSortedLowerLeft[0], new PointF(0, pieceImgBw.Height));
possibleCornersSortedLowerLeft = possibleCornersSortedLowerLeft.Where(c => Utils.Distance(c, new PointF(0, pieceImgBw.Height)) < minCornerDistLowerLeft * PieceFindCornersMaxCornerDistRatio).ToList();
// Combine all possibleCorners from the four lists and discard all combination with too bad angle differences
List <FindCornerRectangleScore> scores = new List <FindCornerRectangleScore>();
for (int indexUpperLeft = 0; indexUpperLeft < possibleCornersSortedUpperLeft.Count; indexUpperLeft++)
{
for (int indexUpperRight = 0; indexUpperRight < possibleCornersSortedUpperRight.Count; indexUpperRight++)
{
for (int indexLowerRight = 0; indexLowerRight < possibleCornersSortedLowerRight.Count; indexLowerRight++)
{
for (int indexLowerLeft = 0; indexLowerLeft < possibleCornersSortedLowerLeft.Count; indexLowerLeft++)
{
if (PluginFactory.CancelToken.IsCancellationRequested)
{
PluginFactory.CancelToken.ThrowIfCancellationRequested();
}
// Possible corner combination
Point[] tmpCorners = new Point[]
{
possibleCornersSortedUpperLeft[indexUpperLeft], // the corners are ordered beginning in the upper left corner and going counter clock wise
possibleCornersSortedLowerLeft[indexLowerLeft],
possibleCornersSortedLowerRight[indexLowerRight],
possibleCornersSortedUpperRight[indexUpperRight]
};
double angleDiff = RectangleDifferenceAngle(tmpCorners);
if (angleDiff > PieceFindCornersMaxAngleDiff)
{
continue;
}
double area = CvInvoke.ContourArea(new VectorOfPoint(tmpCorners));
FindCornerRectangleScore score = new FindCornerRectangleScore()
{
AngleDiff = angleDiff, RectangleArea = area, PossibleCorners = tmpCorners
};
scores.Add(score);
}
}
}
}
// Order the scores by rectangle area (biggest first) and take the PossibleCorners of the biggest rectangle as corners
scores = scores.OrderByDescending(s => s.RectangleArea).ToList();
if (scores.Count > 0)
{
corners.AddRange(scores[0].PossibleCorners);
}
}
if (corners.Count != 4)
{
PluginFactory.LogHandle.Report(new LogEventError(pieceID + " Failed to find correct number of corners. " + corners.Count + " found."));
}
if (PluginFactory.GetGeneralSettingsPlugin().SolverShowDebugResults)
{
using (Image <Rgb, byte> imgCorners = new Image <Rgb, byte>(pieceImgColor))
{
Features2DToolbox.DrawKeypoints(imgCorners, new VectorOfKeyPoint(keyPoints), imgCorners, new Bgr(0, 0, 255)); // Draw the dominant key points
for (int i = 0; i < corners.Count; i++)
{
CvInvoke.Circle(imgCorners, Point.Round(corners[i]), 4, new MCvScalar(0, Math.Max(255 - i * 50, 50), 0), 3);
}
PluginFactory.LogHandle.Report(new LogEventImage(pieceID + " Corners", imgCorners.Bitmap));
imgCorners.Dispose();
}
}
return(corners);
}
public CircleF FindCircle(Image <Gray, Byte> image, int estimatedRadius, int patternType, int error = 30)
{
circles.Clear();
Image <Gray, Byte> bilateralFilteredImage, edgeDetectedImage, eroded, img;
img = image.Clone();
bilateralFilteredImage = new Mat().ToImage <Gray, byte>();
edgeDetectedImage = new Mat().ToImage <Gray, byte>();
eroded = new Mat().ToImage <Gray, byte>();
Mat hierarchy = new Mat();
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint();
//Mat
CvInvoke.MorphologyEx(img, img, MorphOp.Close, GenerateEllipseKernel(13), new Point(-1, -1), 1, BorderType.Default, new MCvScalar());
CvInvoke.BilateralFilter(img, bilateralFilteredImage, 9, 30, 30);
CvInvoke.Canny(bilateralFilteredImage, edgeDetectedImage, 25, 25);
CvInvoke.MorphologyEx(edgeDetectedImage, eroded, MorphOp.Close, GenerateEllipseKernel(11), new Point(-1, -1), 1, BorderType.Default, new MCvScalar());
CvInvoke.FindContours(eroded, contours, hierarchy, RetrType.List, ChainApproxMethod.ChainApproxSimple);
for (int i = 0; i < contours.Size; i++)
{
using (VectorOfPoint contour = contours[i])
{
Rectangle r = CvInvoke.BoundingRectangle(contour);
double w, h;
if (IsSquare(r.Width, r.Height))
{
w = r.Width;
h = r.Height;
double rect_area = ((w * w) / 4) * Math.PI;
CircleF circle = CvInvoke.MinEnclosingCircle(contour);
double circle_area = circle.Radius * circle.Radius * Math.PI;
if ((Math.Abs(rect_area - circle_area) < rect_area / 10) &&
(Math.Abs(Math.Sqrt(circle_area / 3.14) - estimatedRadius) < error) && (w > 21) && (h > 21))
{
CircleFWithScore temp = new CircleFWithScore(circle, CvInvoke.ContourArea(contour) / circle.Area);
circles.Add(temp);
}
}
}
}
//CvInvoke.MatchTemplate(img,templ:templ,)
//CvInvoke.Imshow("edge", eroded);
//var watch = System.Diagnostics.Stopwatch.StartNew();
CircleF result = FindHighestScoreCircle();
if (MatchPattern(image, result, patternType))
{
//watch.Stop();
//var elapsedMs = watch.ElapsedMilliseconds;
//Console.WriteLine("\nFinished pattern matching in " + elapsedMs + "ms");
return(result);
}
else
{
//watch.Stop();
//var elapsedMs = watch.ElapsedMilliseconds;
//Console.WriteLine("\nFinished pattern matching in " + elapsedMs + "ms");
throw new IndexOutOfRangeException();
}
}
public Bitmap FormatImage(Bitmap bitmap)
{
if (bitmap.Width > bitmap.Height)
{
bitmap.RotateFlip(RotateFlipType.Rotate90FlipNone);
}
int originalWidth = bitmap.Width;
int originalHeight = bitmap.Height;
Image <Bgr, Byte> img =
new Image <Bgr, byte>(bitmap).Resize(400, 400, Inter.Linear, true); //resizing is needed for better rectangle detection
int resizedWidth = img.Width;
int resizedHeight = img.Height;
//Convert the image to grayscale and filter out the noise
UMat uimage = new UMat();
CvInvoke.CvtColor(img, uimage, ColorConversion.Bgr2Gray);
//use image pyr to remove noise
UMat pyrDown = new UMat();
CvInvoke.PyrDown(uimage, pyrDown);
CvInvoke.PyrUp(pyrDown, uimage);
// These values work best
double cannyThreshold = 180.0;
double cannyThresholdLinking = 120.0;
UMat cannyEdges = new UMat();
CvInvoke.Canny(uimage, cannyEdges, cannyThreshold, cannyThresholdLinking);
List <Bitmap> cropedImagesList = new List <Bitmap>();
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(cannyEdges, contours, null, RetrType.List, ChainApproxMethod.ChainApproxSimple);
int count = contours.Size;
for (int i = 0; i < count; i++)
{
using (VectorOfPoint contour = contours[i])
using (VectorOfPoint approxContour = new VectorOfPoint())
{
CvInvoke.ApproxPolyDP(contour, approxContour, CvInvoke.ArcLength(contour, true) * 0.05, true);
if (CvInvoke.ContourArea(approxContour, false) > (resizedHeight * resizedWidth) / 3) //only consider contours with area greater than the third of the whole image
{
if (approxContour.Size == 4) //The contour has 4 vertices.
{
//determine if all the angles in the contour are within [70, 110] degree
bool isRectangle = true;
Point[] pts = approxContour.ToArray();
LineSegment2D[] edges = PointCollection.PolyLine(pts, true);
for (int j = 0; j < edges.Length; j++)
{
double angle = Math.Abs(
edges[(j + 1) % edges.Length].GetExteriorAngleDegree(edges[j]));
if (angle < 70 || angle > 110) // these values mean that the angle must be a right angle
{
isRectangle = false;
break;
}
}
if (isRectangle)
{
double[] corners = new double[8];
for (int j = 0; j < 4; j++)
{
corners[2 * j] = Convert.ToDouble(approxContour[j].X) * originalWidth / resizedWidth;
corners[2 * j + 1] = Convert.ToDouble(approxContour[j].Y) * originalHeight / resizedHeight;
}
//crop only if X1 is to the left of X2
if (corners[0] <= corners[2])
{
cropedImagesList.Add(Crop(bitmap, corners));
}
}
}
}
}
}
}
if (FilterCropedImages(cropedImagesList) != null) //if we crop something
{
//crop image and add filter
var result = FilterCropedImages(cropedImagesList);
result = BradleyLocalThreshold(result);
if (result.Width > result.Height)
{
result.RotateFlip(RotateFlipType.Rotate90FlipNone);
return(result);
}
return(result);
}
else
{
//add only filter on original image
var result = BradleyLocalThreshold(bitmap);
return(result);
}
}
public void DetectShapes()
{
StringBuilder msgBuilder = new StringBuilder("Performance: ");
double cannyThreshold = 180.0;
//Load the image from file and resize it for display
var fileImage = _SavePath + _CaptureCounter.ToString() + ".png";
Mat image = CvInvoke.Imread(_SavePath + (_CaptureCounter - 1).ToString() + ".png", Emgu.CV.CvEnum.LoadImageType.AnyColor);
//Convert the image to grayscale and filter out the noise
UMat uimage = new UMat();
CvInvoke.CvtColor(image, uimage, ColorConversion.Bgr2Gray);
//use image pyr to remove noise
UMat pyrDown = new UMat();
CvInvoke.PyrDown(uimage, pyrDown);
CvInvoke.PyrUp(pyrDown, uimage);
//Image<Gray, Byte> gray = img.Convert<Gray, Byte>().PyrDown().PyrUp();
Stopwatch watch = Stopwatch.StartNew();
#region circle detection
#endregion
#region Canny and edge detection
//watch.Reset(); watch.Start();
double cannyThresholdLinking = 120.0;
UMat cannyEdges = new UMat();
CvInvoke.Canny(uimage, cannyEdges, cannyThreshold, cannyThresholdLinking);
#endregion
#region Find triangles and rectangles
watch.Reset(); watch.Start();
List <Triangle2DF> triangleList = new List <Triangle2DF>();
List <Emgu.CV.Structure.RotatedRect> boxList = new List <Emgu.CV.Structure.RotatedRect>(); //a box is a rotated rectangle
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(cannyEdges, contours, null, RetrType.List, ChainApproxMethod.ChainApproxSimple);
int count = contours.Size;
for (int i = 0; i < count; i++)
{
using (VectorOfPoint contour = contours[i])
using (VectorOfPoint approxContour = new VectorOfPoint())
{
CvInvoke.ApproxPolyDP(contour, approxContour, CvInvoke.ArcLength(contour, true) * 0.05, true);
if (CvInvoke.ContourArea(approxContour, false) > 250) //only consider contours with area greater than 250
{
if (approxContour.Size == 4) //The contour has 4 vertices.
{
#region determine if all the angles in the contour are within [80, 100] degree
bool isRectangle = true;
System.Drawing.Point[] pts = approxContour.ToArray();
LineSegment2D[] edges = PointCollection.PolyLine(pts, true);
for (int j = 0; j < edges.Length; j++)
{
double angle = Math.Abs(
edges[(j + 1) % edges.Length].GetExteriorAngleDegree(edges[j]));
//if (angle < 80 || angle > 100)
//{
isRectangle = false;
m_MyAudioSource.Play(0);
SceneManager.LoadScene("Acertou");
//}
}
#endregion
if (isRectangle)
{
boxList.Add(CvInvoke.MinAreaRect(approxContour));
}
}
}
}
}
}
watch.Stop();
#endregion
}
private void button1_Click(object sender, EventArgs e)
{
DialogResult drChosenFile;
OpenFileDialog opf = new OpenFileDialog();
//drChosenFile = ofdOpenFile.ShowDialog(); // open file dialog
/* if (drChosenFile != DialogResult.OK || ofdOpenFile.FileName == "")
* { // if user chose Cancel or filename is blank . . .
* lblChosenFile.Text = "file not chosen"; // show error message on label
* return; // and exit function
* }*/
Mat imgTrainingNumbers;
try
{
imgTrainingNumbers = CvInvoke.Imread("E:\\Emgu\\EmguC\\EmguC\\training_chars.png", LoadImageType.AnyColor);
}
catch (Exception ex)
{ // if error occurred
label1.Text = "unable to open image, error: " + ex.Message; // show error message on label
return; // and exit function
}
if (imgTrainingNumbers == null)
{ // if image could not be opened
label1.Text = "unable to open image"; // show error message on label
return; // and exit function
}
label1.Text = opf.FileName; //update label with file name
Mat imgGrayscale = new Mat();
Mat imgBlurred = new Mat(); // declare various images
Mat imgThresh = new Mat();
Mat imgThreshCopy = new Mat();
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint();
//Matrix<Single> mtxClassifications = new Matrix<Single>();
//Matrix<Single> mtxTrainingImages = new Matrix<Single>();
Mat matTrainingImagesAsFlattenedFloats = new Mat();
//possible chars we are interested in are digits 0 through 9 and capital letters A through Z, put these in list intValidChars
var intValidChars = new List <int>(new int[] {
'0', '1', '2', '3', '4', '5', '6', '7', '8', '9',
'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J',
'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T',
'U', 'V', 'W', 'X', 'Y', 'Z'
});
CvInvoke.CvtColor(imgTrainingNumbers, imgGrayscale, ColorConversion.Bgr2Gray); //convert to grayscale
CvInvoke.GaussianBlur(imgGrayscale, imgBlurred, new Size(5, 5), 0); //blur
//threshold image from grayscale to black and white
CvInvoke.AdaptiveThreshold(imgBlurred, imgThresh, 255.0, AdaptiveThresholdType.GaussianC, ThresholdType.BinaryInv, 11, 2);
CvInvoke.Imshow("imgThresh", imgThresh); //show threshold image for reference
imgThreshCopy = imgThresh.Clone(); //make a copy of the thresh image, this in necessary b/c findContours modifies the image
//get external countours only
CvInvoke.FindContours(imgThreshCopy, contours, null, RetrType.External, ChainApproxMethod.ChainApproxSimple);
int intNumberOfTrainingSamples = contours.Size;
Matrix <Single> mtxClassifications = new Matrix <Single>(intNumberOfTrainingSamples, 1); //this is our classifications data structure
//this is our training images data structure, note we will have to perform some conversions to write to this later
Matrix <Single> mtxTrainingImages = new Matrix <Single>(intNumberOfTrainingSamples, RESIZED_IMAGE_WIDTH * RESIZED_IMAGE_HEIGHT);
//this keeps track of which row we are on in both classifications and training images,
int intTrainingDataRowToAdd = 0; //note that each sample will correspond to one row in
//both the classifications XML file and the training images XML file
for (int i = 0; i <= contours.Size - 1; ++i)
{ //for each contour
if (CvInvoke.ContourArea(contours[i]) > MIN_CONTOUR_AREA)
{ //if contour is big enough to consider
Rectangle boundingRect = CvInvoke.BoundingRectangle(contours[i]); //get the bounding rect
CvInvoke.Rectangle(imgTrainingNumbers, boundingRect, new MCvScalar(0.0, 0.0, 255.0), 2); //draw red rectangle around each contour as we ask user for input
Mat imgROItoBeCloned = new Mat(imgThresh, boundingRect); //get ROI image of current char
Mat imgROI = imgROItoBeCloned.Clone(); //make a copy so we do not change the ROI area of the original image
Mat imgROIResized = new Mat();
//resize image, this is necessary for recognition and storage
CvInvoke.Resize(imgROI, imgROIResized, new Size(RESIZED_IMAGE_WIDTH, RESIZED_IMAGE_HEIGHT));
CvInvoke.Imshow("imgROI", imgROI); //show ROI image for reference
CvInvoke.Imshow("imgROIResized", imgROIResized); //show resized ROI image for reference
CvInvoke.Imshow("imgTrainingNumbers", imgTrainingNumbers); //show training numbers image, this will now have red rectangles drawn on it
int intChar = CvInvoke.WaitKey(0); //get key press
if (intChar == 27)
{ //if esc key was pressed
CvInvoke.DestroyAllWindows();
return; //exit the function
}
else if (intValidChars.Contains(intChar))
{ //else if the char is in the list of chars we are looking for . . .
mtxClassifications[intTrainingDataRowToAdd, 0] = Convert.ToSingle(intChar); //write classification char to classifications Matrix
//now add the training image (some conversion is necessary first) . . .
//note that we have to covert the images to Matrix(Of Single) type, this is necessary to pass into the KNearest object call to train
Matrix <Single> mtxTemp = new Matrix <Single>(imgROIResized.Size);
Matrix <Single> mtxTempReshaped = new Matrix <Single>(1, RESIZED_IMAGE_WIDTH * RESIZED_IMAGE_HEIGHT);
imgROIResized.ConvertTo(mtxTemp, DepthType.Cv32F); //convert Image to a Matrix of Singles with the same dimensions
for (int intRow = 0; intRow <= RESIZED_IMAGE_HEIGHT - 1; ++intRow)
{ //flatten Matrix into one row by RESIZED_IMAGE_WIDTH * RESIZED_IMAGE_HEIGHT number of columns
for (int intCol = 0; intCol <= RESIZED_IMAGE_WIDTH - 1; ++intCol)
{
mtxTempReshaped[0, (intRow * RESIZED_IMAGE_WIDTH) + intCol] = mtxTemp[intRow, intCol];
}
}
for (int intCol = 0; intCol <= (RESIZED_IMAGE_WIDTH * RESIZED_IMAGE_HEIGHT) - 1; ++intCol)
{ //write flattened Matrix into one row of training images Matrix
mtxTrainingImages[intTrainingDataRowToAdd, intCol] = mtxTempReshaped[0, intCol];
}
intTrainingDataRowToAdd = intTrainingDataRowToAdd + 1; //increment which row, i.e. sample we are on
}
}
}
label1.Text = label1.Text + "training complete !!" + "\n" + "\n";
//save classifications to file
XmlSerializer xmlSerializer = new XmlSerializer(mtxClassifications.GetType());
StreamWriter streamWriter;
try
{
streamWriter = new StreamWriter("classifications.xml"); //attempt to open classifications file
}
catch (Exception ex)
{ //if error is encountered, show error and return
label1.Text = "\n" + label1.Text + "unable to open 'classifications.xml', error:" + "\n";
label1.Text = label1.Text + ex.Message + "\n" + "\n";
return;
}
xmlSerializer.Serialize(streamWriter, mtxClassifications);
streamWriter.Close();
//save training images to file
xmlSerializer = new XmlSerializer(mtxTrainingImages.GetType());
try
{
streamWriter = new StreamWriter("images.xml"); // attempt to open images file
}
catch (Exception ex)
{ // if error is encountered, show error and return
label1.Text = "\n" + label1.Text + "unable to open 'images.xml', error:" + "\n";
label1.Text = label1.Text + ex.Message + "\n" + "\n";
return;
}
xmlSerializer.Serialize(streamWriter, mtxTrainingImages);
streamWriter.Close();
label1.Text = "\n" + label1.Text + "file writing done" + "\n";
MessageBox.Show("Training complete, file writing done !!");
}
public void PerformShapeDetection()
{
if (textBox1.Text != String.Empty)
{
StringBuilder msgBuilder = new StringBuilder("Performance: ");
//Load the image from file and resize it for display
Image <Bgr, Byte> img =
new Image <Bgr, byte>(textBox1.Text)
.Resize(400, 400, Emgu.CV.CvEnum.Inter.Linear, true);
//Convert the image to grayscale and filter out the noise
UMat uimage = new UMat();
CvInvoke.CvtColor(img, uimage, ColorConversion.Bgr2Gray);
//use image pyr to remove noise
UMat pyrDown = new UMat();
CvInvoke.PyrDown(uimage, pyrDown);
CvInvoke.PyrUp(pyrDown, uimage);
//Image<Gray, Byte> gray = img.Convert<Gray, Byte>().PyrDown().PyrUp();
#region circle detection
Stopwatch watch = Stopwatch.StartNew();
double cannyThreshold = 180.0;
double circleAccumulatorThreshold = 120;
CircleF[] circles = CvInvoke.HoughCircles(uimage, HoughType.Gradient, 2.0, 20.0, cannyThreshold, circleAccumulatorThreshold, 5);
watch.Stop();
msgBuilder.Append(String.Format("Hough circles - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
#region Canny and edge detection
watch.Reset(); watch.Start();
double cannyThresholdLinking = 120.0;
UMat cannyEdges = new UMat();
CvInvoke.Canny(uimage, cannyEdges, cannyThreshold, cannyThresholdLinking);
LineSegment2D[] lines = CvInvoke.HoughLinesP(
cannyEdges,
1, //Distance resolution in pixel-related units
Math.PI / 45.0, //Angle resolution measured in radians.
20, //threshold
30, //min Line width
10); //gap between lines
watch.Stop();
msgBuilder.Append(String.Format("Canny & Hough lines - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
#region Find triangles and rectangles
watch.Reset(); watch.Start();
List <Triangle2DF> triangleList = new List <Triangle2DF>();
List <RotatedRect> boxList = new List <RotatedRect>(); //a box is a rotated rectangle
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(cannyEdges, contours, null, RetrType.List, ChainApproxMethod.ChainApproxSimple);
int count = contours.Size;
for (int i = 0; i < count; i++)
{
using (VectorOfPoint contour = contours[i])
using (VectorOfPoint approxContour = new VectorOfPoint())
{
CvInvoke.ApproxPolyDP(contour, approxContour, CvInvoke.ArcLength(contour, true) * 0.05, true);
if (CvInvoke.ContourArea(approxContour, false) > 250) //only consider contours with area greater than 250
{
if (approxContour.Size == 3) //The contour has 3 vertices, it is a triangle
{
Point[] pts = approxContour.ToArray();
triangleList.Add(new Triangle2DF(
pts[0],
pts[1],
pts[2]
));
}
else if (approxContour.Size == 4) //The contour has 4 vertices.
{
#region determine if all the angles in the contour are within [80, 100] degree
bool isRectangle = true;
Point[] pts = approxContour.ToArray();
LineSegment2D[] edges = PointCollection.PolyLine(pts, true);
for (int j = 0; j < edges.Length; j++)
{
double angle = Math.Abs(
edges[(j + 1) % edges.Length].GetExteriorAngleDegree(edges[j]));
if (angle < 80 || angle > 100)
{
isRectangle = false;
break;
}
}
#endregion
if (isRectangle)
{
boxList.Add(CvInvoke.MinAreaRect(approxContour));
}
}
}
}
}
}
watch.Stop();
msgBuilder.Append(String.Format("Triangles & Rectangles - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
OriginalImageBox.Image = img;
this.Text = msgBuilder.ToString();
#region draw triangles and rectangles
Mat triangleRectangleImage = new Mat(img.Size, DepthType.Cv8U, 3);
triangleRectangleImage.SetTo(new MCvScalar(0));
foreach (Triangle2DF triangle in triangleList)
{
CvInvoke.Polylines(triangleRectangleImage, Array.ConvertAll(triangle.GetVertices(), Point.Round), true, new Bgr(Color.DarkBlue).MCvScalar, 2);
}
foreach (RotatedRect box in boxList)
{
CvInvoke.Polylines(triangleRectangleImage, Array.ConvertAll(box.GetVertices(), Point.Round), true, new Bgr(Color.DarkOrange).MCvScalar, 2);
}
TriangleandRectangleImageBox.Image = triangleRectangleImage;
#endregion
#region draw circles
Mat circleImage = new Mat(img.Size, DepthType.Cv8U, 3);
circleImage.SetTo(new MCvScalar(0));
foreach (CircleF circle in circles)
{
CvInvoke.Circle(circleImage, Point.Round(circle.Center), (int)circle.Radius, new Bgr(Color.Brown).MCvScalar, 2);
}
CircleImageBox.Image = circleImage;
#endregion
#region draw lines
Mat lineImage = new Mat(img.Size, DepthType.Cv8U, 3);
lineImage.SetTo(new MCvScalar(0));
foreach (LineSegment2D line in lines)
{
CvInvoke.Line(lineImage, line.P1, line.P2, new Bgr(Color.Green).MCvScalar, 2);
}
LineImageBox.Image = lineImage;
#endregion
}
}
// Update is called once per frame
void Update()
{
imageOrig = webCam.QueryFrame();
if (imageOrig != null)
{
imageGray = new Mat();
imageHSV = new Mat();
Mat imageAverage = new Mat();
Mat imageMedian = new Mat();
Mat imageGaussian = new Mat();
Mat imageBilateral = new Mat();
imageResult = new Mat();
CvInvoke.Flip(imageOrig, imageOrig, FlipType.Horizontal);
CvInvoke.Resize(imageOrig, imageOrig, new Size(imSize, imSize * webCam.Height / webCam.Width));
CvInvoke.CvtColor(imageOrig, imageGray, ColorConversion.Bgr2Gray);
CvInvoke.CvtColor(imageOrig, imageHSV, ColorConversion.Bgr2Hsv);
// Draw Original Image
//CvInvoke.Imshow(imNameOrig, imageOrig);
// Draw Original Image
CvInvoke.Imshow(imNameColor, imageHSV);
Mat filteredImg = new Mat();
//CvInvoke.BilateralFilter( imageHSV, filteredImg, 3, 75, 75 );
//CvInvoke.GaussianBlur( imageHSV, filteredImg, new Size(7,7), 0 );
CvInvoke.MedianBlur(imageHSV, filteredImg, 7);
Image <Hsv, System.Byte> rangeImg = filteredImg.ToImage <Hsv, System.Byte>();
// Yellow 40-70; 0-255; 0-255
Hsv bottomHsv = new Hsv(18, 127, 127); // 0-179, 0-255, 0-255
Hsv topHsv = new Hsv(35, 240, 240);
if (detectColorBlue)
{
// BLue 230-180; 0-255; 0-255
bottomHsv = new Hsv(80, 70, 70); // 0-179, 0-255, 0-255
topHsv = new Hsv(120, 250, 250);
}
//rangeImg.Data[0,0,0];
Mat imagBW = rangeImg.InRange(bottomHsv, topHsv).Mat;
int elementSize = 5;
Mat structureElement = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(elementSize, elementSize), new Point(-1, -1));
CvInvoke.Erode(imagBW, imagBW, structureElement, new Point(-1, -1), 1, BorderType.Constant, new MCvScalar(0));
CvInvoke.Dilate(imagBW, imagBW, structureElement, new Point(-1, -1), 1, BorderType.Constant, new MCvScalar(0));
// Contours
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint();
int biggestContourIndex = -1;
double biggestContourArea = -1;
Mat hierarchy = new Mat();
CvInvoke.FindContours(imagBW, contours, hierarchy, RetrType.List, ChainApproxMethod.ChainApproxNone);
if (contours.Size > 0)
{
biggestContourIndex = 0;
biggestContourArea = CvInvoke.ContourArea(contours[biggestContourIndex]);
}
for (int i = 1; i < contours.Size; i++)
{
double currentArea = CvInvoke.ContourArea(contours[i]);
if (currentArea > biggestContourArea)
{
biggestContourIndex = i;
biggestContourArea = currentArea;
}
}
if (contours.Size > 0)
{
VectorOfPoint biggestContour = contours[biggestContourIndex];
CvInvoke.DrawContours(imageOrig, contours, biggestContourIndex, new MCvScalar(255, 0, 0), 5);
if (previousBiggestContourArea > 0) // Object entering
{
if (biggestContourArea > previousBiggestContourArea * 1.6)
{
// Going Foward
//Debug.Log( "Front" );
if (canon)
{
canon.Shoot();
}
}
else if (biggestContourArea * 1.6 < previousBiggestContourArea)
{
// Going backward
//Debug.Log( "Back" );
}
}
previousBiggestContourArea = biggestContourArea;
//* Centroid
MCvMoments moment = CvInvoke.Moments(contours[biggestContourIndex]);
int cx = (int)(moment.M10 / moment.M00);
int cy = (int)(moment.M01 / moment.M00);
imageOrig = drawPoint(imageOrig, cx, cy, 5);
if (canon)
{
canon.setHorizontalPosition(cx / (float)imSize);
}
//*/
//* Top Point
Point top = biggestContour[0];
for (int i = 1; i < biggestContour.Size; i++)
{
Point p = biggestContour[i];
if (top.Y > p.Y)
{
top = p;
}
}
if (canon)
{
canon.setVerticalPosition(((float)imageOrig.SizeOfDimemsion[0] - top.Y) / (float)imageOrig.SizeOfDimemsion[0]);
}
imageOrig = drawPoint(imageOrig, top.X, top.Y, 5, 255, 0, 0);
//*/
}
else
{
// Object leaving
previousBiggestContourArea = -1;
}
CvInvoke.Imshow("BW", imagBW);
CvInvoke.Imshow(imNameOrig, imageOrig);
// Filtering
/*CvInvoke.Blur( imageHSV, imageAverage, new Size(5,5), new Point(-1,-1) );
* CvInvoke.Imshow("Average", imageAverage);
* CvInvoke.MedianBlur( imageHSV, imageMedian, 5 );
* CvInvoke.Imshow("Median", imageMedian);
* CvInvoke.GaussianBlur( imageHSV, imageGaussian, new Size(5,5), 0 );
* CvInvoke.Imshow("Gaussian", imageGaussian);
* CvInvoke.BilateralFilter( imageHSV, imageBilateral, 3, 75, 75 );
* CvInvoke.Imshow("Bilateral", imageBilateral);*/
//CvInvoke.Imshow(imNameResult, imageResult);
// Storing
writer.Write(imageOrig);
}
else
{
webCam = new VideoCapture(imAddress);
}
if (Input.GetKeyDown(KeyCode.Escape))
{
Application.Quit();
}
}
private void button1_Click(object sender, EventArgs e)
{
//단일쓰레드 일때 문자열을 많이 읽고 변경할경우 좋은 StringBuilder 사용 ( string 비해 속도 매우빠름)
// https://blog.naver.com/impressives2/221338797755 참고
StringBuilder msgBuilder = new StringBuilder("Performance: ");
//Image<Bgr, Byte> img =
// new Image<Bgr, byte>(fileNameTextBox.Text)
// .Resize(400, 400, Emgu.CV.CvEnum.Inter.Linear, true);
// OpenFileDialog 으로 image를 가지고와야 하므로 객체생성
OpenFileDialog ofd = new OpenFileDialog();
//// OpenFileDialog 상자가 열리고 확인버튼을 눌렀을 때 !!
if (ofd.ShowDialog() == DialogResult.OK)
{
//https://dic1224.blog.me/220841161411 Resize는 확대축소
//https://blog.naver.com/PostView.nhn?blogId=dic1224&logNo=220841171866&parentCategoryNo=&categoryNo=152&viewDate=&isShowPopularPosts=true&from=search
//위에꺼는 소스와 그림(?)
//https://dic1224.blog.me/220841161411 Emgu.CV.CvEnum.Inter.Linear의 구조
img = new Image <Bgr, Byte>(ofd.FileName).Resize(400, 400, Emgu.CV.CvEnum.Inter.Linear, true);
}
UMat uimage = new UMat();
// img 객체 이미지가 Bgr 형식으로 되어있으니 Bgr2Gray로 그레이시킨 후 uimage에 값을 출력
CvInvoke.CvtColor(img, uimage, ColorConversion.Bgr2Gray);
//use image pyr to remove noise
UMat pyrDown = new UMat();
CvInvoke.PyrDown(uimage, pyrDown); // 노이즈제거 및 그레이된걸 샘플링다운으로 추출 (출력..?)
CvInvoke.PyrUp(pyrDown, uimage); // 노이즈제거 후 샘플링업으로 추출( 출력..)
Image <Gray, Byte> gray = img.Convert <Gray, Byte>().PyrDown().PyrUp();
#region circle detection
// 경과시간을 정확히 추출하는 객체 하나생성 ( 0초로) 시간을 재야함
Stopwatch watch = Stopwatch.StartNew();
//원형 만들기 위한 수치
double cannyThreshold = 180.0;
double circleAccumulatorThreshold = 120;
//uimage는 노이즈제거, 그레이, 샘플링다운상태
//uimage로 원그리기
CircleF[] circles = CvInvoke.HoughCircles(uimage, HoughType.Gradient, 2.0, 20.0, cannyThreshold, circleAccumulatorThreshold, 5);
// 원다그렸으니 시간을 멈추기
watch.Stop();
// 얼마나 걸렸는지 출력
msgBuilder.Append(String.Format("Hough circles - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
#region Canny and edge detection
// 0초로만들고 다시시작
watch.Reset(); watch.Start();
double cannyThresholdLinking = 120.0;
UMat cannyEdges = new UMat();
//Canny알고리즘사용하여 cannyEdges 객체에 값넣기 3,4번째 임계값1,2
CvInvoke.Canny(uimage, cannyEdges, cannyThreshold, cannyThresholdLinking);
//cannyEdges 한 후 Hough lines한다
LineSegment2D[] lines = CvInvoke.HoughLinesP(
cannyEdges,
1, //Distance resolution in pixel-related units
Math.PI / 45.0, //Angle resolution measured in radians.
20, //threshold
30, //min Line width
10); //gap between lines
// 측정을멈춤
watch.Stop();
//몇초 걸렸는지 출력
msgBuilder.Append(String.Format("Canny & Hough lines - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
#region Find triangles and rectangles
// 새로운 테스트위해 시간을 0초로 되돌리고 시작
watch.Reset(); watch.Start();
// triangles 객체생성
List <Triangle2DF> triangleList = new List <Triangle2DF>();
// rectangles 객체생성
List <RotatedRect> boxList = new List <RotatedRect>(); //a box is a rotated rectangle
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(cannyEdges, contours, null, RetrType.List, ChainApproxMethod.ChainApproxSimple);
int count = contours.Size;
for (int i = 0; i < count; i++)
{
using (VectorOfPoint contour = contours[i])
using (VectorOfPoint approxContour = new VectorOfPoint())
{
CvInvoke.ApproxPolyDP(contour, approxContour, CvInvoke.ArcLength(contour, true) * 0.05, true);
if (CvInvoke.ContourArea(approxContour, false) > 100) //only consider contours with area greater than 250
{
if (approxContour.Size == 3) //The contour has 3 vertices, it is a triangle
{
Point[] pts = approxContour.ToArray();
triangleList.Add(new Triangle2DF(
pts[0],
pts[1],
pts[2]
));
}
else if (approxContour.Size == 4) //The contour has 4 vertices.
{
#region determine if all the angles in the contour are within [80, 100] degree
bool isRectangle = true;
Point[] pts = approxContour.ToArray();
LineSegment2D[] edges = PointCollection.PolyLine(pts, true);
for (int j = 0; j < edges.Length; j++)
{
double angle = Math.Abs(
edges[(j + 1) % edges.Length].GetExteriorAngleDegree(edges[j]));
if (angle < 90 || angle > 110)
{
isRectangle = false;
break;
}
}
#endregion
if (isRectangle)
{
boxList.Add(CvInvoke.MinAreaRect(approxContour));
}
}
}
}
}
}
//측정종료
watch.Stop();
//출력
msgBuilder.Append(String.Format("Triangles & Rectangles - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
// 원본 img는 그대로 pictureBox에 출력
originalImageBox.Image = img;
// 폼제목을 측정한 데이터로 변경
this.Text = msgBuilder.ToString();
#region draw triangles and rectangles
Image <Bgr, Byte> triangleRectangleImage = img.CopyBlank();
foreach (Triangle2DF triangle in triangleList)
{
triangleRectangleImage.Draw(triangle, new Bgr(Color.DarkBlue), 2);
}
foreach (RotatedRect box in boxList)
{
triangleRectangleImage.Draw(box, new Bgr(Color.DarkOrange), 2);
}
// triangles and rectangles 한 img를 pictureBox2에출력
triangleRectangleImageBox.Image = triangleRectangleImage;
#endregion
Image <Bgr, Byte> circleImage = img.CopyBlank();
#region draw circles img.CopyBlank();
foreach (CircleF circle in circles)
{
circleImage.Draw(circle, new Bgr(Color.Brown), 2);
}
// circles 한 img를 pictureBox3에출력
circleImageBox.Image = circleImage;
#endregion
#region draw lines
Image <Bgr, Byte> lineImage = img.CopyBlank();
foreach (LineSegment2D line in lines)
{
lineImage.Draw(line, new Bgr(Color.Green), 2);
}
// Detected Lines 한 이미지를 pictureBox4에출력
lineImageBox.Image = lineImage;
#endregion
}
public void GetBoundries(Image <Gray, Byte> binaryBackground, out List <Point[]> boundries, out List <Point[]> artefacts, out List <RotatedRect> boxes)
{
//Find outer boundries
double minimumContourArea = 250;
double minimumBoundryArea = 1000;
//double approximationFactor = 0.001;
List <Point[]> allBoundries = new List <Point[]>();
List <Point[]> allObjects = new List <Point[]>();
List <RotatedRect> boxList = new List <RotatedRect>();
using (Image <Gray, Byte> filteredBinary = binaryBackground.SmoothMedian(7))
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
using (Mat hierarchy = new Mat())
{
CvInvoke.FindContours(filteredBinary, contours, hierarchy, RetrType.Tree, ChainApproxMethod.ChainApproxNone);
var temp = hierarchy.ToImage <Bgra, Byte>();
int count = contours.Size;
List <int> boundryIds = new List <int>();
for (int i = 0; i < count; i++)
{
using (VectorOfPoint contour = contours[i])
{
double contourArea = CvInvoke.ContourArea(contour);
if (contourArea >= minimumBoundryArea)
{
Bgra currentContour = temp[0, i];
if (currentContour.Alpha == 0)
{
allBoundries.Add(contour.ToArray());
boundryIds.Add(i);
}
}
}
}
for (int i = 0; i < count; i++)
{
using (VectorOfPoint contour = contours[i])
{
double contourArea = CvInvoke.ContourArea(contour);
if (contourArea >= minimumContourArea)
{
Bgra currentContour = temp[0, i];
if (!boundryIds.Contains(i) && boundryIds.Contains((int)currentContour.Alpha))
{
bool isRectangle = true;
bool isCircle = false;
//Can the object be approximated as a circle or rectangle?
using (VectorOfPoint apxContour = new VectorOfPoint())
{
double epsilon = CvInvoke.ArcLength(contour, true) * 0.05;
CvInvoke.ApproxPolyDP(contour, apxContour, epsilon, true);
if (apxContour.Size == 4) //The contour has 4 vertices.
{
Point[] pts = apxContour.ToArray();
LineSegment2D[] edges = PointCollection.PolyLine(pts, true);
for (int j = 0; j < edges.Length; j++)
{
double angle = Math.Abs(edges[(j + 1) % edges.Length].GetExteriorAngleDegree(edges[j]));
if (angle < 70 || angle > 110)
{
isRectangle = false;
break;
}
}
if (isRectangle)
{
boxList.Add(CvInvoke.MinAreaRect(apxContour));
}
}
else
{
isRectangle = false;
}
}
if (!isRectangle && !isCircle)
{
allObjects.Add(contour.ToArray());
}
}
}
}
}
}
//Find mouse
//mousePoints = null;
//using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
//{
// CvInvoke.FindContours(binaryMouse, contours, null, RetrType.External, ChainApproxMethod.ChainApproxNone);
// int count = contours.Size;
// double maxArea = 0;
// for (int j = 0; j < count; j++)
// {
// using (VectorOfPoint contour = contours[j])
// {
// double contourArea = CvInvoke.ContourArea(contour);
// if (contourArea >= maxArea)
// {
// maxArea = contourArea;
// mousePoints = contour.ToArray();
// }
// }
// }
//}
boundries = allBoundries;
artefacts = allObjects;
boxes = boxList;
//Check if any contours can be approximated as shapes
//We now have a list of boundries, if there's more than one it means something is sticking across the screen
if (allBoundries.Count > 1)
{
//Need to find points from all boundries that are effectively parallel
}
//Image<Bgr, Byte> allContourImage = FirstFrame.Clone();
//allContourImage.DrawPolyline(mousePoints, true, new Bgr(Color.Yellow), 2);
//allContourImage.DrawPolyline(allBoundries.ToArray(), true, new Bgr(Color.Red), 2);
//allContourImage.DrawPolyline(allObjects.ToArray(), true, new Bgr(Color.LightGreen), 2);
//foreach (var box in boxList)
//{
// allContourImage.Draw(box.GetVertices().Select(x => new Point((int)x.X, (int)x.Y)).ToArray(), new Bgr(Color.Aqua), 2);
//}
}
public void FindLicensePlateAndCharacters(VectorOfVectorOfPoint contours, int[,] hierachy, int idx, Mat gray, Mat canny, List <string> licenses)
{
for (; idx >= 0; idx = hierachy[idx, 0])
{
int numberOfChildren = GetNumberOfChildren(hierachy, idx);
//if it does not contains any children (charactor), it is not a license plate region
if (numberOfChildren == 0)
{
continue;
}
using (VectorOfPoint contour = contours[idx])
{
if (CvInvoke.ContourArea(contour) > 200)
{
if (numberOfChildren < 3)
{
// If the contour has less than 3 children, it is not a license plate (assuming license plate has at least 3 charactor)
// However we should search the children of this contour to see if any of them is a license plate
FindLicensePlateAndCharacters(contours, hierachy, hierachy[idx, 2], gray, canny, licenses);
continue;
}
RotatedRect box = CvInvoke.MinAreaRect(contour);
if (box.Angle < -45.0)
{
float tmp = box.Size.Width;
box.Size.Width = box.Size.Height;
box.Size.Height = tmp;
box.Angle += 90.0f;
}
else if (box.Angle > 45.0)
{
float tmp = box.Size.Width;
box.Size.Width = box.Size.Height;
box.Size.Height = tmp;
box.Angle -= 90.0f;
}
double whRatio = (double)box.Size.Width / box.Size.Height;
if (!(3.0 < whRatio && whRatio < 10.0))
//if (!(1.0 < whRatio && whRatio < 2.0))
{
//if the width height ratio is not in the specific range,it is not a license plate
//However we should search the children of this contour to see if any of them is a license plate
//Contour<Point> child = contours.VNext;
if (hierachy[idx, 2] > 0)
{
FindLicensePlateAndCharacters(contours, hierachy, hierachy[idx, 2], gray, canny, licenses);
}
continue;
}
using (UMat tmp1 = new UMat())
using (UMat tmp2 = new UMat())
{
PointF[] srcCorners = box.GetVertices();
PointF[] destCorners = new PointF[] {
new PointF(0, box.Size.Height - 1),
new PointF(0, 0),
new PointF(box.Size.Width - 1, 0),
new PointF(box.Size.Width - 1, box.Size.Height - 1)
};
using (Mat rot = CameraCalibration.GetAffineTransform(srcCorners, destCorners))
CvInvoke.WarpAffine(gray, tmp1, rot, Size.Round(box.Size));
//resize the license plate such that the front is ~ 10-12. This size of front results in better accuracy from tesseract
Size approxSize = new Size(240, 180);
double scale = Math.Min(approxSize.Width / box.Size.Width, approxSize.Height / box.Size.Height);
Size newSize = new Size((int)Math.Round(box.Size.Width * scale), (int)Math.Round(box.Size.Height * scale));
CvInvoke.Resize(tmp1, tmp2, newSize, 0, 0, Inter.Cubic);
//removes some pixels from the edge
int edgePixelSize = 2;
Rectangle newRoi = new Rectangle(new Point(edgePixelSize, edgePixelSize),
tmp2.Size - new Size(2 * edgePixelSize, 2 * edgePixelSize));
UMat plate = new UMat(tmp2, newRoi);
List <UMat> filteredCharaters = FilterPlate.GetCharacters(plate);
_plateViewModel.filteredCharatersInSinglePlate.Add(filteredCharaters);
_plateViewModel.DetectedPlates.Add(plate.Bitmap.ToBitmapImage());
foreach (var character in filteredCharaters)
{
_plateViewModel.FilteredDetectedCharacters.Add(character.Bitmap.ToBitmapImage());
}
}
}
}
}
}
// 填充缺陷轮廓
private Image <Bgr, byte> ContourFilling3(Image <Bgr, byte> pic)
{
Image <Bgr, byte> outpic = new Image <Bgr, byte>(pic.Size);
Image <Ycc, byte> ycc = pic.Convert <Ycc, byte>();
for (int i = 0; i < ycc.Height; i++)
{
for (int j = 0; j < ycc.Width; j++)
{
if (ycc[i, j].Cr > 35 && ycc[i, j].Cr < 148 &&
ycc[i, j].Cb > 48 && ycc[i, j].Cb < 141)
{
ycc[i, j] = new Ycc(0, 0, 0);
}
else
{
ycc[i, j] = new Ycc(255, 255, 255);
}
}
}
Image <Gray, byte> gray = ycc.Convert <Gray, byte>();
gray = gray.ThresholdBinary(new Gray(100), new Gray(255));
gray = gray.Canny(100, 60);
Image <Gray, byte> outcon = new Image <Gray, byte>(pic.Size);
VectorOfVectorOfPoint con = new VectorOfVectorOfPoint();
CvInvoke.FindContours(gray, con, outcon, RetrType.External, ChainApproxMethod.ChainApproxNone);
int n = 0;
for (int i = 0; i < con.Size; i++)
{
if (CvInvoke.ContourArea(con[i]) > 0)
{
n++;
}
}
textBox1.Text = "共" + n.ToString() + "个缺陷" + " " + "\n";
n = 0;
for (int i = 0; i < con.Size; i++)
{
if (CvInvoke.ContourArea(con[i]) > 0)
{
CvInvoke.DrawContours(outpic, con, i, new MCvScalar(0, 255, 0), 5);
textBox1.Text = textBox1.Text + "第" + (++n).ToString() + "个缺陷的面积为" + CvInvoke.ContourArea(con[i]) + " \n";
}
}
CvInvoke.AddWeighted(outpic, 0.5, picture, 0.5, 0, outpic);
return(outpic);
}
private void Btn_anwserReg_Click(object sender, EventArgs e)
{
if (this.ib_middleCut.Image == null)
{
MessageBox.Show("裁剪图片不能为空");
return;
}
CVHelper commonUse = new CVHelper();
Mat src = new Image <Bgr, byte>(ib_middleCut.Image.Bitmap).Mat;// new Mat();
//CvInvoke.PyrMeanShiftFiltering(src1, src, 25, 10, 1, new MCvTermCriteria(5, 1));
//commonUse.SaveMat(src, "降噪后");
//commonUse.SaveMat(src1, "降噪后原始");
Mat dst = new Mat();
Mat src_gray = new Mat();
CvInvoke.CvtColor(src, src_gray, Emgu.CV.CvEnum.ColorConversion.Bgr2Gray);
//存储灰度图片
commonUse.SaveMat(src_gray, "灰度");
#region 二值化
//二值化
Mat mat_threshold = new Mat();
int myThreshold = Convert.ToInt32(num_threshold.Value);
CvInvoke.Threshold(src_gray, mat_threshold, myThreshold, 255, Emgu.CV.CvEnum.ThresholdType.BinaryInv);
commonUse.SaveMat(mat_threshold, "二值化");
//思路 close -腐蚀-腐蚀-膨胀
//形态学膨胀
Mat mat_dilate = commonUse.MyDilate(mat_threshold, Emgu.CV.CvEnum.MorphOp.Close);
commonUse.SaveMat(mat_dilate, "形态学膨胀");
//mat_dilate = commonUse.MyDilate(mat_dilate, Emgu.CV.CvEnum.MorphOp.Close);
//commonUse.SaveMat(mat_dilate, "形态学膨胀1");
mat_dilate = commonUse.MyDilate(mat_dilate, Emgu.CV.CvEnum.MorphOp.Erode);
commonUse.SaveMat(mat_dilate, "形态学膨胀腐蚀1");
mat_dilate = commonUse.MyDilate(mat_dilate, Emgu.CV.CvEnum.MorphOp.Erode);
commonUse.SaveMat(mat_dilate, "形态学膨胀腐蚀2");
mat_dilate = commonUse.MyDilate(mat_dilate, Emgu.CV.CvEnum.MorphOp.Dilate);
commonUse.SaveMat(mat_dilate, "形态学膨胀2");
#endregion
//边缘检测
CvInvoke.Canny(mat_dilate, dst, Convert.ToInt32(this.num_Min.Value), Convert.ToInt32(this.num_Max.Value), Convert.ToInt32(this.num_apertureSize.Value));
commonUse.SaveMat(dst, "边缘检测");
//寻找答题卡矩形边界（所有的矩形）
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint(); //创建VectorOfVectorOfPoint数据类型用于存储轮廓
VectorOfVectorOfPoint validContours = new VectorOfVectorOfPoint(); //有效的，所有的选项的
CvInvoke.FindContours(dst, contours, null, Emgu.CV.CvEnum.RetrType.Ccomp,
Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxSimple, new Point(8, 8));//提取轮廓
//打印所以后矩形面积和周长
int size = contours.Size;
for (int i = 0; i < size; i++)
{
var item = contours[i];
var tempArea = CvInvoke.ContourArea(item);
var tempArc = CvInvoke.ArcLength(item, true);
Console.WriteLine($"面积：{tempArea}；周长：{tempArc}");;
if (tempArea > 200 && tempArea < 2000)
{
validContours.Push(item);
}
}
//CvInvoke.ApproxPolyDP
//画出所有轮廓
Mat middleMat = new Image <Bgr, byte>(this.ib_middleCut.Image.Bitmap).Mat;
CvInvoke.DrawContours(middleMat, validContours, -1, new MCvScalar(0, 0, 255), 1);
this.ib_result.Image = middleMat;
commonUse.SaveMat(middleMat, "画出所有轮廓的");
//画出所有矩形
Mat tmpMat = new Image <Bgr, byte>(this.ib_middleCut.Image.Bitmap).Mat;
List <Rectangle> rectangles = commonUse.GetRectList(validContours, false);
rectangles.ForEach(rect =>
{
CvInvoke.Rectangle(tmpMat, rect, new MCvScalar(0, 0, 255));
});
commonUse.SaveMat(tmpMat, "画出所有矩形轮廓的");
this.ib_result.Image = tmpMat;
}
FileOperation fileOperation = new FileOperation();//实例化处理文本的类
/// <summary>
/// 获取各区块的轮廓
/// </summary>
public void getContours(TextBox txtTypeName, PictureBox ptb) //找最近的轮廓
{
GLB.Match_success = false; //重新检测赋值
Image <Gray, byte> dnc = new Image <Gray, byte>(GLB.BUFW, GLB.BUFH);
Image <Gray, byte> threshImage = new Image <Gray, byte>(GLB.BUFW, GLB.BUFH);
CvInvoke.CvtColor(GLB.frame, threshImage, ColorConversion.Bgra2Gray);//灰度化
//CvInvoke.BilateralFilter(threshImage, threshImage, 10, 10, 4);//双边滤波
//CvInvoke.GaussianBlur(threshImage, threshImage, new Size(3, 3), 4);//高斯滤波
CvInvoke.BoxFilter(threshImage, threshImage, Emgu.CV.CvEnum.DepthType.Cv8U, new Size(3, 3), new Point(-1, -1));//方框滤波
#region
//var kernal1 = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(3, 3), new Point(-1, -1));
//CvInvoke.Dilate(threshImage, threshImage, kernal1, new Point(-1, -1), 2, BorderType.Default, new MCvScalar());//膨胀
//var kernal1 = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(3, 3), new Point(-1, -1));
//CvInvoke.Erode(threshImage, threshImage, kernal1, new Point(-1, -1), 2, BorderType.Default, new MCvScalar());//腐蚀
//方式1
//CvInvoke.Threshold(threshImage, threshImage, 100, 255, ThresholdType.BinaryInv | ThresholdType.Otsu);//二值化
//if (Mainform.runMode == 6)//匹配托盘
//{
// var kernal1 = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(9, 9), new Point(-1, -1));
// CvInvoke.Erode(threshImage, threshImage, kernal1, new Point(-1, -1), 1, BorderType.Default, new MCvScalar());//腐蚀
//}
//else//匹配箱子
//{
// var kernal1 = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(3, 3), new Point(-1, -1));
// CvInvoke.Erode(threshImage, threshImage, kernal1, new Point(-1, -1), 2, BorderType.Default, new MCvScalar());//腐蚀
//}
//方式2
//if (Mainform.runMode == 6)//匹配托盘
//{
// var kernal1 = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(9, 9), new Point(-1, -1));
// CvInvoke.Dilate(threshImage, threshImage, kernal1, new Point(-1, -1), 1, BorderType.Default, new MCvScalar());//膨胀
//}
//else //加了膨胀跳动更大
//{
// var kernal1 = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(5, 5), new Point(-1, -1));
// CvInvoke.Dilate(threshImage, threshImage, kernal1, new Point(-1, -1), 1, BorderType.Default, new MCvScalar());//膨胀
//}
//ptb.Image = threshImage.ToBitmap();
#endregion
//检测连通域，每一个连通域以一系列的点表示，FindContours方法只能得到第一个域:
try
{
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint(2000); //区块集合
CvInvoke.FindContours(threshImage, contours, dnc, RetrType.Ccomp, ChainApproxMethod.ChainApproxSimple); //轮廓集合
GLB.block_num = 0;
Dictionary <int, VectorOfPoint> mycontours = new Dictionary <int, VectorOfPoint>(100);//序号，轮廓
mycontours.Clear();
for (int k = 0; k < contours.Size; k++)
{
double area = CvInvoke.ContourArea(contours[k]); //获取各连通域的面积
if (area > 100000 && area < 800000) //根据面积作筛选(指定最小面积,最大面积):
{
if (!mycontours.ContainsKey(k))
{
mycontours.Add(k, contours[k]);
}
}
}
float my_depth_temp = GLB.myp3d[(GLB.BUFH / 2 * GLB.BUFW + GLB.BUFW / 2) * 3 + 2];
if (mycontours.Count == 0 && Mainform.ProduceArrive == true && Mainform.CarryMode == 0 && Mainform.runMode == 1 && (my_depth_temp > 1400 || double.IsNaN(my_depth_temp)))//空车来，小车自动离开
{
Mainform.ProduceArrive = false;
Mainform.SetCarryArrive(0); //修改产品没送到
ArrayList array = new ArrayList(); //多条SQL语句数组
string sql = "update Agv_list set isworking =0,stowerid ='',pronum =0 where agvid in(select agvid from Agvmission_list where fstatus =7 and messionType =1 and stowerid='" + GLB.RobotId + "')"; //修改小车状态
string sql1 = "update Agvmission_list set fstatus =6 ,actionenddate=getdate() where fstatus =7 and messionType =1 and stowerid='" + GLB.RobotId + "'"; //修改任务 等待状态为完成状态
array.Add(sql);
array.Add(sql1);
bool isok = MyDataLib.transactionOp_list(array);
Mainform.SetRobotStatus(2, "等待送货");//修改码垛机器人状态
}
//按面积最大排序 生成新的字典
Dictionary <int, VectorOfPoint> mycontours_SortedByKey = new Dictionary <int, VectorOfPoint>(100);//序号，轮廓;
mycontours_SortedByKey.Clear();
mycontours_SortedByKey = mycontours.OrderByDescending(o => CvInvoke.ContourArea(o.Value)).ToDictionary(p => p.Key, o => o.Value);
GLB.obj.Clear();
foreach (int k in mycontours_SortedByKey.Keys)
{
OBJ obj = new OBJ();
{
if (!GLB.obj.ContainsKey(GLB.block_num))
{
GLB.obj.Add(GLB.block_num, obj); //不含这个，就添加
}
GLB.obj[GLB.block_num].typName = txtTypeName.Text.Replace(" ", ""); // 对象名称
if (getMinAreaRect(mycontours_SortedByKey[k], GLB.block_num) == true) //获取最小外接矩形并处理相关参数
{
if (GLB.img_mode == 0) //匹配模式
{
if (Device_Macth(GLB.block_num) == true) //与库对比，生成工件位置，法向量,旋转角
{
Thread.Sleep(400);
break;
}
}
}
GLB.TitleStr += "block_num=" + GLB.block_num;
GLB.block_num++;//区块计数器
}
}
}
catch (Exception ex)
{
Console.WriteLine("发生错误： " + ex.Message);
throw;
}
}
/// <summary>
/// Detect basic elements (such as circlr, line, rectangle and triangle)
/// </summary>
/// <param name="argPath"></param>
/// <param name="argtMode"></param>
/// <returns></returns>
public static DetectBasicEleementResult DetectBasicElement(string argPath, DetectMode argtMode)
{
StringBuilder msgBuilder = new StringBuilder("Performance: ");
//Load the image from file and resize it for display
Image <Bgr, byte> img = new Image <Bgr, byte>(argPath).Resize(400, 400, Emgu.CV.CvEnum.Inter.Linear, true);
//Convert the image to grayscale and filter out the noise
UMat uimage = new UMat();
CvInvoke.CvtColor(img, uimage, ColorConversion.Bgr2Gray);
//use image pyr to remove noise
UMat pyrDown = new UMat();
CvInvoke.PyrDown(uimage, pyrDown);
CvInvoke.PyrUp(pyrDown, uimage);
//Image<Gray, Byte> gray = img.Convert<Gray, Byte>().PyrDown().PyrUp();
#region circle detection
CircleF[] circles = null;
Stopwatch watch = new Stopwatch();
double cannyThreshold = 180.0;
if (argtMode == DetectMode.IncludeCircle)
{
watch = Stopwatch.StartNew();
double circleAccumulatorThreshold = 120;
circles = CvInvoke.HoughCircles(uimage, HoughType.Gradient, 2.0, 20.0, cannyThreshold, circleAccumulatorThreshold, 5);
watch.Stop();
msgBuilder.Append(String.Format("Hough circles - {0} ms; ", watch.ElapsedMilliseconds));
}
#endregion
#region Canny and edge detection
watch.Reset(); watch.Start();
double cannyThresholdLinking = 120.0;
UMat cannyEdges = new UMat();
CvInvoke.Canny(uimage, cannyEdges, cannyThreshold, cannyThresholdLinking);
LineSegment2D[] lines = CvInvoke.HoughLinesP(cannyEdges,
1, //Distance resolution in pixel-related units
Math.PI / 45.0, //Angle resolution measured in radians.
20, //threshold
30, //min Line width
10); //gap between lines
watch.Stop();
msgBuilder.Append(String.Format("Canny & Hough lines - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
#region Find triangles and rectangles
watch.Reset(); watch.Start();
List <Triangle2DF> triangleList = new List <Triangle2DF>();
List <RotatedRect> boxList = new List <RotatedRect>(); //a box is a rotated rectangle
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(cannyEdges, contours, null, RetrType.List, ChainApproxMethod.ChainApproxSimple);
int count = contours.Size;
for (int i = 0; i < count; i++)
{
using (VectorOfPoint contour = contours[i])
using (VectorOfPoint approxContour = new VectorOfPoint())
{
CvInvoke.ApproxPolyDP(contour, approxContour, CvInvoke.ArcLength(contour, true) * 0.05, true);
if (CvInvoke.ContourArea(approxContour, false) > 250) //only consider contours with area greater than 250
{
if (approxContour.Size == 3) //The contour has 3 vertices, it is a triangle
{
Point[] pts = approxContour.ToArray();
triangleList.Add(new Triangle2DF(pts[0], pts[1], pts[2]));
}
else if (approxContour.Size == 4) //The contour has 4 vertices.
{
#region determine if all the angles in the contour are within [80, 100] degree
bool isRectangle = true;
Point[] pts = approxContour.ToArray();
LineSegment2D[] edges = PointCollection.PolyLine(pts, true);
for (int j = 0; j < edges.Length; j++)
{
double angle = Math.Abs(edges[(j + 1) % edges.Length].GetExteriorAngleDegree(edges[j]));
if (angle < 80 || angle > 100)
{
isRectangle = false;
break;
}
}
#endregion
if (isRectangle)
{
boxList.Add(CvInvoke.MinAreaRect(approxContour));
}
}
}
}
}
}
watch.Stop();
msgBuilder.Append(String.Format("Triangles & Rectangles - {0} ms; ", watch.ElapsedMilliseconds));
#endregion
return(new DetectBasicEleementResult(img, triangleList, boxList, circles, lines, msgBuilder.ToString()));
}
private void button2_Click(object sender, EventArgs e) //procesiranje slike
{
c = 0; fields = new List <Image <Gray, byte> >(); grid = new int[9, 9];
OpenFileDialog opf = new OpenFileDialog();
if (opf.ShowDialog() != DialogResult.OK)
{
return;
}
Image <Gray, Byte> gray = new Image <Gray, byte>(opf.FileName);
imageBox1.Image = gray.Clone();
Image <Gray, Byte> izhod = new Image <Gray, byte>(gray.Width, gray.Height);
// binariziranje(adaptivna pragovna segmentacija)
izhod = gray.ThresholdAdaptive(new Gray(255), Emgu.CV.CvEnum.AdaptiveThresholdType.MeanC, Emgu.CV.CvEnum.ThresholdType.BinaryInv, 11, new Gray(11));
izhod._SmoothGaussian(1); //glajenje, (ni obvezno potrebno)
// imageBox2.Image = izhod.Clone();
var nova = izhod.Clone();
//posopek za iskanje najvecjega povezanega objekta
VectorOfVectorOfPoint vvp = new VectorOfVectorOfPoint();
Mat hierarchy = new Mat();
CvInvoke.FindContours(izhod, vvp, hierarchy, Emgu.CV.CvEnum.RetrType.Tree, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxNone);
int largest_contour_index = 0;
double largest_area = 0;
VectorOfPoint largestContour;
for (int i = 0; i < vvp.Size; i++)
{
double a = CvInvoke.ContourArea(vvp[i], false);
if (a > largest_area)
{
largest_area = a;
largest_contour_index = i;
}
}
largestContour = vvp[largest_contour_index];
Point[] lc = largestContour.ToArray();
//iskanje kote za perspektivna transformacija
Point topleft = new Point(gray.Width, gray.Height);
Point topright = new Point(0, gray.Height);
Point botright = new Point(0, 0);
Point botleft = new Point(gray.Width, 0);
foreach (Point p in lc)
{
if ((p.X + p.Y) < (topleft.X + topleft.Y))
{
topleft = p;
}
else if ((p.X - p.Y) > (topright.X - topright.Y))
{
topright = p;
}
else if ((p.X + p.Y) > (botright.X + botright.Y))
{
botright = p;
}
else if ((p.Y - p.X) > (botleft.Y - botleft.X))
{
botleft = p;
}
}
//prerisemo gridlines, da se znebimo linije, potem ostanejo le stevilke(lazja razpoznava)
CvInvoke.DrawContours(nova, vvp, largest_contour_index, new MCvScalar(0, 0, 0), 6, Emgu.CV.CvEnum.LineType.EightConnected, hierarchy, 1);
Image <Gray, Byte> warp = new Image <Gray, byte>(450, 450);
PointF[] src = new PointF[] { topleft, topright, botright, botleft };
PointF[] dst = new PointF[] { new Point(0, 0), new Point(450, 0), new Point(450, 450), new Point(0, 450) };
Mat warpmat = CvInvoke.GetPerspectiveTransform(src, dst); //racunamo matrika za transformacija
CvInvoke.WarpPerspective(nova, warp, warpmat, new Size(450, 450)); //izvedemo transformacija
//imageBox1.Image = nova;
imageBox2.Image = warp;
//warp._Erode(1); //krcenje ali sirjenje, ni potrebno
//warp._Dilate(1);
//razpoznava stevilk, 2 moznosti (izbira so radiogumbov)
if (radioButton1.Checked)
{
tess = new Tesseract(@"C:/Emgu/emgucv-windows-universal 3.0.0.2157/bin/", null, OcrEngineMode.Default, "123456789 ");
}
fields = new List <Image <Gray, byte> >(); //hranim polja za lazjo debagiranje
for (int i = 0; i < 9; i++)
{
for (int j = 0; j < 9; j++)
{
Image <Gray, Byte> temp = (warp.GetSubRect(new Rectangle(j * 50 + 3, i * 50 + 3, 44, 44))).Clone(); //malo izpustimo po robu polja
temp._SmoothGaussian(1);
Gray sum = temp.GetSum(); //ce ni dovolj beli pikslov(dele objektov), ni stevilka.
if (sum.Intensity < 30000)
{
continue;
}
//spet iscemo najvecji element v polju, predvidevam da je stevilo
VectorOfVectorOfPoint vvptemp = new VectorOfVectorOfPoint();
Mat hierarchytemp = new Mat();
CvInvoke.FindContours(temp, vvptemp, hierarchytemp, Emgu.CV.CvEnum.RetrType.Tree, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxNone);
int ind = 0;
double area = 100;
VectorOfPoint contour;
for (int k = 0; k < vvptemp.Size; k++)
{
double ar = CvInvoke.ContourArea(vvptemp[k], false);
if (ar > area)
{
area = ar;
ind = k;
}
}
if (area == 100)
{
continue; //ce je najvecjega kontura area manj kot 100 (polje je 44x44), povzamemem da ni stevilo
}
contour = vvptemp[ind]; //kontura stevlike
var tempimg = new Image <Gray, Byte>(44, 44, new Gray(0));
CvInvoke.DrawContours(tempimg, vvptemp, ind, new MCvScalar(255, 0, 0), -1, Emgu.CV.CvEnum.LineType.EightConnected, hierarchytemp);
//narisemo notranjosti najvecjega kontura v novi sliki z belo barvo
fields.Add(tempimg); //dodamo za pogled
if (radioButton2.Checked)
{
Rectangle br = CvInvoke.BoundingRectangle(contour);
int indeks = 0;
double vrednost = double.MaxValue;
for (int q = 0; q < 9; q++)
{ //racunamo podobnost s vsakem vzorcu
var kraj = tempimg.GetSubRect(new Rectangle(br.X, br.Y, vzorci[q].Width, vzorci[q].Height));
var pod = vzorci[q].AbsDiff(kraj);
var podobnost = pod.GetSum();
if (podobnost.Intensity < vrednost)
{
indeks = q + 1; //ker je zero based
vrednost = podobnost.Intensity;
}
}
grid[i, j] = indeks;//najbolj podobni je zaznana stevilka
}
else
{
tess.Recognize(tempimg); //raspoznava slike s pomocjo tesseract OCR vgrajen v openCV
var x = tess.GetCharacters();
if (x.Length == 1)
{
grid[i, j] = Convert.ToInt32(x[0].Text);
}
}
}
}
NarisiStevilke(); //izpisemo stevilki v polje
}
private List <Result> DetectBanknotesTrain(Mat image, float minimumMatchAllowed = 0.07f, float minimuTargetAreaPercentage = 0.05f, float maxDistanceRatio = 0.75f, float reprojectionThresholPercentage = 0.01f,
double confidence = 0.99, int maxIters = 5000, int minimumNumerInliers = 8)
{
object locker = new object();
List <Result> detectorResults = new List <Result>();
MKeyPoint[] mKeyPoints;
SIFT sift = new SIFT();
mKeyPoints = sift.Detect(image);
VectorOfKeyPoint keypointsEvalImage = new VectorOfKeyPoint();
keypointsEvalImage.Push(mKeyPoints);
if (keypointsEvalImage.Size < 4)
{
return(detectorResults);
}
Mat descriptorsEvalImage = new Mat();
sift.Compute(image, keypointsEvalImage, descriptorsEvalImage);
Features2DToolbox.DrawKeypoints(image, keypointsEvalImage, image, new Bgr(0, 0, 255), Features2DToolbox.KeypointDrawType.Default);
float bestMatch = 0;
Result bestDetectorResult = new Result();
int trainDetectorsSize = DetectedBanknotes.Count;
bool validDetection = true;
float reprojectionThreshold = image.Cols * reprojectionThresholPercentage;
do
{
bestMatch = 0;
Parallel.For(0, trainDetectorsSize, i =>
{
DetectedBanknotes[(int)i].UpdateCurrentLODIndex(ref image, 0.6999999881F);
Result detectorResult = DetectedBanknotes[(int)i].AnalyzeImageEval(ref keypointsEvalImage, ref descriptorsEvalImage, maxDistanceRatio, reprojectionThreshold, confidence, maxIters, minimumNumerInliers);
if (detectorResult.GetBestROIMatch() > minimumMatchAllowed)
{
float contourArea = (float)CvInvoke.ContourArea(detectorResult.GetTrainContour());
float imageArea = (float)(image.Cols * image.Rows);
float contourAreaPercentage = contourArea / imageArea;
if (contourAreaPercentage > minimuTargetAreaPercentage)
{
double contourAspectRatio = _util.ComputeContourAspectRatio(detectorResult.GetTrainContour());
if (contourAspectRatio > _contourAspectRatioRange.X && contourAspectRatio < _contourAspectRatioRange.Y)
{
double contourCircularity = _util.ComputeContourCircularity(detectorResult.GetTrainContour());
if (contourCircularity > _contourCircularityRange.X && contourCircularity < _contourCircularityRange.Y)
{
if (CvInvoke.IsContourConvex(detectorResult.GetTrainContour()))
{
lock (locker)
{
if (detectorResult.GetBestROIMatch() > bestMatch)
{
bestMatch = detectorResult.GetBestROIMatch();
bestDetectorResult = detectorResult;
}
}
}
}
}
}
}
});
validDetection = bestMatch > minimumMatchAllowed && bestDetectorResult.GetInliers().Size > minimumNumerInliers;
if (bestDetectorResult != null && validDetection)
{
detectorResults.Add(bestDetectorResult);
_util.RemoveInliersFromKeypointsAndDescriptors(bestDetectorResult.GetInliers(), ref keypointsEvalImage, ref descriptorsEvalImage);
}
} while (validDetection);
return(detectorResults);
}
private void Processing(Image <Bgr, byte> ImgSource, TextColor Color = TextColor.White)
{
Rectangle ROICode = new Rectangle();
mImgDetected = ImgSource.Copy();
// create and ROI image
Rectangle ROI = new Rectangle(ImgSource.Width / 2, ImgSource.Height / 10, ImgSource.Width, ImgSource.Height / 4);
mImgDetected.ROI = ROI;
// filter noise
//detect code
using (Image <Gray, byte> imgGray = mImgDetected.Convert <Gray, byte>())
{
using (Image <Gray, byte> imgFilter = new Image <Gray, byte>(imgGray.Size))
{
CvInvoke.BilateralFilter(imgGray, imgFilter, 9, 49, 49);
using (Mat k = CvInvoke.GetStructuringElement(Emgu.CV.CvEnum.ElementShape.Rectangle, new Size(5, 5), new Point(-1, -1)))
{
if (Color == TextColor.White)
{
CvInvoke.Erode(imgFilter, imgFilter, k, new Point(-1, -1), 1, Emgu.CV.CvEnum.BorderType.Default, new MCvScalar());
}
else
{
CvInvoke.Dilate(imgFilter, imgFilter, k, new Point(-1, -1), 1, Emgu.CV.CvEnum.BorderType.Default, new MCvScalar());
}
}
using (Image <Gray, double> ImgSobel = new Image <Gray, double>(imgFilter.Size))
{
CvInvoke.Sobel(imgFilter, ImgSobel, Emgu.CV.CvEnum.DepthType.Cv64F, 1, 0, kSize: 1);
CvInvoke.ConvertScaleAbs(ImgSobel, imgFilter, 2, 0);
CvInvoke.Threshold(imgFilter, imgFilter, 20, 255, Emgu.CV.CvEnum.ThresholdType.Binary);
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(imgFilter, contours, null, Emgu.CV.CvEnum.RetrType.External, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxSimple);
for (int i = 0; i < contours.Size; i++)
{
double s = CvInvoke.ContourArea(contours[i]);
Rectangle bound = CvInvoke.BoundingRectangle(contours[i]);
if (bound.Height > 65 || s < 10)
{
CvInvoke.DrawContours(imgFilter, contours, i, new MCvScalar(0), -1);
}
}
}
using (Mat k = CvInvoke.GetStructuringElement(Emgu.CV.CvEnum.ElementShape.Rectangle, new Size(107, 1), new Point(-1, -1)))
{
CvInvoke.MorphologyEx(imgFilter, imgFilter, Emgu.CV.CvEnum.MorphOp.Close, k, new Point(-1, -1), 1, Emgu.CV.CvEnum.BorderType.Default, new MCvScalar());
}
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(imgFilter, contours, null, Emgu.CV.CvEnum.RetrType.External, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxSimple);
double large_area = 0;
int index_large = 0;
for (int i = 0; i < contours.Size; i++)
{
double s = CvInvoke.ContourArea(contours[i]);
if (large_area < s)
{
large_area = s;
index_large = i;
}
}
Rectangle boxFirstLine = CvInvoke.BoundingRectangle(contours[index_large]);
Rectangle boxSecondLine = new Rectangle();
for (int i = 0; i < contours.Size; i++)
{
Rectangle b = CvInvoke.BoundingRectangle(contours[i]);
if (b.Y - boxFirstLine.Y < 120 && b.Y - boxFirstLine.Y > 0 && b.Width > 30)
{
boxSecondLine = CvInvoke.BoundingRectangle(contours[i]);
break;
}
}
ROICode = new Rectangle(boxFirstLine.X - 20, boxFirstLine.Y - 20, boxFirstLine.Width + 40, boxSecondLine.Y + boxSecondLine.Height + 60 - boxFirstLine.X);
ROICode.X = ROICode.X < 0 ? 0: ROICode.X;
ROICode.Y = ROICode.Y < 0 ? 0 : ROICode.Y;
ROICode.Width = ROICode.X + ROICode.Width > mImgDetected.Width ? mImgDetected.Width - ROICode.X : ROICode.Width;
ROICode.Height = ROICode.Y + ROICode.Height > mImgDetected.Height ? mImgDetected.Height - ROICode.Y : ROICode.Height;
mImgCroped = mImgDetected.Copy();
mImgCroped.ROI = ROICode;
CvInvoke.Rectangle(mImgDetected, ROICode, new MCvScalar(255, 0, 0), 3);
mImgDetected.ROI = new Rectangle();
imb3.Image = mImgCroped.Bitmap;
}
}
}
}
// segment char text
mImgSegment = new Image <Gray, byte>(mImgCroped.Size);
mImgCharSegment = mImgCroped.Copy();
CvInvoke.CvtColor(mImgCroped, mImgSegment, Emgu.CV.CvEnum.ColorConversion.Bgr2Gray);
using (Mat k = CvInvoke.GetStructuringElement(Emgu.CV.CvEnum.ElementShape.Rectangle, new Size(5, 5), new Point(-1, -1)))
{
CvInvoke.MorphologyEx(mImgSegment, mImgSegment, Emgu.CV.CvEnum.MorphOp.Open, k, new Point(-1, -1), 1, Emgu.CV.CvEnum.BorderType.Default, new MCvScalar());
}
Image <Gray, byte> img_decode = mImgSegment.Copy();
CvInvoke.BitwiseNot(img_decode, img_decode);
CvInvoke.Imwrite("test.png", img_decode);
CvInvoke.Threshold(mImgSegment, mImgSegment, 127, 255, Emgu.CV.CvEnum.ThresholdType.Binary);
using (Mat k = CvInvoke.GetStructuringElement(Emgu.CV.CvEnum.ElementShape.Rectangle, new Size(3, 3), new Point(-1, -1)))
{
CvInvoke.MorphologyEx(mImgSegment, mImgSegment, Emgu.CV.CvEnum.MorphOp.Open, k, new Point(-1, -1), 1, Emgu.CV.CvEnum.BorderType.Default, new MCvScalar());
}
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(mImgSegment, contours, null, Emgu.CV.CvEnum.RetrType.External, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxSimple);
for (int i = 0; i < contours.Size; i++)
{
Rectangle bound = CvInvoke.BoundingRectangle(contours[i]);
if (bound.Height > 60 || bound.Height < 30 || bound.Width > 35)
{
CvInvoke.DrawContours(mImgSegment, contours, i, new MCvScalar(0), -1);
}
}
}
using (VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint())
{
CvInvoke.FindContours(mImgSegment, contours, null, Emgu.CV.CvEnum.RetrType.External, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxSimple);
for (int i = 0; i < contours.Size; i++)
{
Rectangle bound = CvInvoke.BoundingRectangle(contours[i]);
CvInvoke.Rectangle(mImgCharSegment, bound, new MCvScalar(0, 255, 0), 2);
}
}
CvInvoke.Threshold(mImgSegment, mImgSegment, 127, 255, Emgu.CV.CvEnum.ThresholdType.BinaryInv);
imb4.Image = mImgSegment.Bitmap;
imb5.Image = mImgCharSegment.Bitmap;
string code = Read(img_decode);
Console.WriteLine(code);
imb2.Image = mImgDetected.Bitmap;
}
private void DetectBall(Mat image, Image <Hsv, byte> img, Hsv seuilBas, Hsv seuilHaut, String suffix)
{
//Left Bin Image
Mat imageBinLeft = img.InRange(seuilBas, seuilHaut).Mat;
int operationSize = 1;
Mat structuringElement = CvInvoke.GetStructuringElement(ElementShape.Ellipse, new Size(2 * operationSize + 1, 2 * operationSize + 1), new Point(operationSize, operationSize));
MCvScalar constante = new MCvScalar(10);
MCvScalar colorConst = new MCvScalar(100, 255, 100);
VectorOfVectorOfPoint contourObject = new VectorOfVectorOfPoint();
int indexBestContour = -1;
CvInvoke.Erode(imageBinLeft, imageBinLeft, structuringElement, new Point(operationSize, operationSize), 3, BorderType.Default, constante);
CvInvoke.Dilate(imageBinLeft, imageBinLeft, structuringElement, new Point(operationSize, operationSize), 3, BorderType.Default, constante);
CvInvoke.FindContours(imageBinLeft, contourObject, null, RetrType.Ccomp, ChainApproxMethod.ChainApproxNone);
for (int i = 0; i < contourObject.Size; i++)
{
if (i == 0)
{
longMaxContour = CvInvoke.ContourArea(contourObject[i]);
indexBestContour = 0;
}
else
{
if (longMaxContour < CvInvoke.ContourArea(contourObject[i]))
{
longMaxContour = CvInvoke.ContourArea(contourObject[i]);
indexBestContour = i;
}
}
}
if (indexBestContour > -1)
{
double area = CvInvoke.ContourArea(contourObject[indexBestContour]);
if (suffix == "Left")
{
leftZ = area / (image.Width * image.Height) * zNormalFactor;
}
else if (suffix == "Right")
{
rightZ = area / (image.Width * image.Height) * zNormalFactor;
}
else
{
throw new NotImplementedException();
}
}
else
{
if (suffix == "Left")
{
leftZ = -1;
}
else if (suffix == "Right")
{
rightZ = -1;
}
else
{
throw new NotImplementedException();
}
}
//CvInvoke.Imshow("Mon Image" + suffix, imageBinLeft);
if (contourObject.Size > 0)
{
CvInvoke.DrawContours(img, contourObject, indexBestContour, colorConst, 2);
}
if (suffix == "Left")
{
RILeftHand.texture = ImageToTexture(imageBinLeft.Clone(), textureLeftHand);
}
else
{
RIRightHand.texture = ImageToTexture(imageBinLeft.Clone(), textureRightHand);
}
}
// Maximum Value, Minimum Value and their locations
// Mean Color or Mean Intensity
public void calcularRegionProps(Image <Gray, byte> inputRegionIMG, double AreaMin)
{
// Declaração do vetor de vetores de pontos
Emgu.CV.Util.VectorOfVectorOfPoint vetordeVetdePontos = new Emgu.CV.Util.VectorOfVectorOfPoint();
// Declaração de uma matriz
Mat hierarquia = new Mat();
// Aplicação da função FindContour
CvInvoke.FindContours(
inputRegionIMG // Recebe a imagem de entrada
, vetordeVetdePontos // Recebe um vetor de pontos de contorno
, hierarquia // Recebe a hierarquia dos pontos
, Emgu.CV.CvEnum.RetrType.Tree // Recebe o tipo de arvore e contornos
, Emgu.CV.CvEnum.ChainApproxMethod.ChainApproxNone // Tip de aproximação aos contornos
, new Point(0, 0) // Offset do ponto, posso omitir ou declarar um ponto a 0 0
);
Image <Bgr, Byte> input = inputRegionIMG.Convert <Bgr, byte>();
//Até aqui encontro o contorno. Deve ser só 1!!!, portanto deve ser o contorno 0,
//mas mesmo assim vamos fazer um teste para ver qual o contorno a usar
// Pontos buffer
PointF buffer_Minx = new PointF(inputRegionIMG.Width, inputRegionIMG.Height);
PointF buffer_MaxX = new PointF(0, 0);
PointF buffer_MinY = new PointF(inputRegionIMG.Width, inputRegionIMG.Height);
PointF buffer_MaxY = new PointF(0, 0);
for (int i = 0; i < vetordeVetdePontos.Size; i++)
{
Area = Math.Abs(CvInvoke.ContourArea(vetordeVetdePontos[i], true)); // calcula a area do contorno
if (Area >= AreaMin)
{
for (int iter = 0; iter < vetordeVetdePontos[i].Size; iter++)
{
//----------------- Calculo do extreme -----------------
// Calcula o valor do ponto mais à esquerda
if (vetordeVetdePontos[i][iter].X < buffer_Minx.X)
{
buffer_Minx = vetordeVetdePontos[i][iter];
}
// Calcula o valor do ponto mais à direita
if (vetordeVetdePontos[i][iter].X > buffer_MaxX.X)
{
buffer_MaxX = vetordeVetdePontos[i][iter];
}
// Calcula o valor do ponto Y mais em cima
if (vetordeVetdePontos[i][iter].Y < buffer_MinY.Y)
{
buffer_MinY = vetordeVetdePontos[i][iter];
}
// Calcula o valor do ponto Y mais em baixo
if (vetordeVetdePontos[i][iter].Y > buffer_MaxY.Y)
{
buffer_MaxY = vetordeVetdePontos[i][iter];
}
//----------------- Fim do calculo do extreme -----------------
}
// ------------- Calculo do Centroid ---------------------
Moments momento = CvInvoke.Moments(vetordeVetdePontos[i]);
int X = (int)(momento.M10 / momento.M00);
int Y = (int)(momento.M01 / momento.M00);
Centroid = new PointF(X, Y);
// ------------------------------------------------------
// ------------ Calculo do AspectRatio ------------------
AspectRatio = inputRegionIMG.Width / inputRegionIMG.Height;
//-------------------------------------------------------
//------------- Calculo da BoundingBox ------------------
BoundingBox = CvInvoke.BoundingRectangle(vetordeVetdePontos[i]);
//-------------------------------------------------------
// ------------ Calculo do Extent -------------------
float rect_area = BoundingBox.Width * BoundingBox.Height;
Extent = (float)Area / rect_area;
// ------------------------------------------------------
// --------------- ConvectHULL --------------------------
CvInvoke.ConvexHull(vetordeVetdePontos[i], ConvexHull, false);
//-------------------------------------------------------
// --------------- ConvectHULL_area ---------------------
ConvexHull_area = CvInvoke.ContourArea(ConvexHull);
//-------------------------------------------------------
//----------------- Solidity ---------------------------
Solidity = Area / ConvexHull_area;
// ------------------------------------------------------
//-------------- Diametro Equivalente -------------------
EquivalentDiameter = Math.Sqrt(4 * Area / Math.PI);
// ------------------------------------------------------
//--------------- Circulo Envolvente --------------------
CirculoEnvolvente = CvInvoke.MinEnclosingCircle(vetordeVetdePontos[i]);
//-------------------------------------------------------
//--------------- Circulo Perimetro --------------------
perimetro = CvInvoke.ArcLength(vetordeVetdePontos[i], true);
// -----------------------------------------------------
// -------------- Circularity (Fator de forma)----------
Circularity = (4 * Math.PI * Area) / (perimetro * perimetro);
//------------------------------------------------------
// --------------- Verifica se é convexo ---------------
isConvex = CvInvoke.IsContourConvex(vetordeVetdePontos[i]);
//------------------------------------------------------
// ------------- Apriximação do contorno ---------------
CvInvoke.ApproxPolyDP(
vetordeVetdePontos[i], // Cada vetor de um contorno iterado
ContourApproximation, // Vetor que vai conter a aproximação
0.1 * perimetro, // Expande o perimetro
true // Calcula um aproximação ao contorno externo
);
// -----------------------------------------------------
// ------------- Devolve o contorno --------------------
Contorno = vetordeVetdePontos[i];
// ------------------------------------------------------
// ------------ Retangulo rodado ---------------------
RotatedRect retanguloRodado = CvInvoke.MinAreaRect(vetordeVetdePontos[i]);
PointF[] vetorPontos = CvInvoke.BoxPoints(retanguloRodado);
BoundingBoxRectRodado = new Point[vetorPontos.Length];
for (int iterador = 0; iterador < vetorPontos.Length; iterador++)
{
BoundingBoxRectRodado[iterador].X = (int)vetorPontos[iterador].X;
BoundingBoxRectRodado[iterador].Y = (int)vetorPontos[iterador].Y;
}
// ------------ AnguloRecExterior ----------------------
AnguloRectExterior = retanguloRodado.Angle;
// -----------------------------------------------------
// ------------ EllipseImagem --------------------------
EllipseValores = CvInvoke.FitEllipseAMS(vetordeVetdePontos[i]);
// -----------------------------------------------------
// Fitting a Line ---------------
//---------------------------
// salta do ciclo for
i = vetordeVetdePontos.Size;
}
}
Extreme.Mais_a_esquerda = buffer_Minx;
Extreme.Mais_a_Direita = buffer_MaxX;
Extreme.Mais_em_baixo = buffer_MaxY;
Extreme.Mais_em_cima = buffer_MinY;
}
private void Capture_ImageGrabbed(object sender, EventArgs e)
{
lock (Capture)
{
if (Busy)
{
return;
}
Busy = true;
}
try
{
Capture.Retrieve(CameraFeed);
var imageFrame = CameraFeed.ToImage <Bgr, byte>();
#region smooth
var imgSmoothed = imageFrame.PyrDown().PyrUp();
imgSmoothed._SmoothGaussian(Smooth);
imageBox2.Image = imgSmoothed;
#endregion
#region Color filter
var imgColorFiltered = imgSmoothed.InRange(new Bgr(Bmin, Gmin, Rmin), new Bgr(Bmax, Gmax, Rmax));
imgColorFiltered = imgColorFiltered.PyrDown().PyrUp();
imgColorFiltered._SmoothGaussian(Smooth); // 3
imageBox3.Image = imgColorFiltered;
#endregion
#region erosion, delete noise
//var imageErode = imgColorFiltered.Erode(2);
//imageBox4.Image = imageErode;
#endregion
#region
//var imageDilate = imageErode.Dilate(5);
//imageBox5.Image = imageDilate;
#endregion
#region opening
// open > erosion and dilate
Mat kernel = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(5, 5), new Point(-1, -1));
var opening = imgColorFiltered.MorphologyEx(MorphOp.Open, kernel, new Point(-1, -1), 1, BorderType.Default, new MCvScalar(1.0));
imageBox4.Image = opening;
#endregion
#region closing
// open > erosion and dilate
//Mat kernel = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(5, 5), new Point(-1, -1));
//var opening = imgColorFiltered.MorphologyEx(MorphOp.Close, kernel, new Point(-1, -1), 1, BorderType.Default, new MCvScalar(1.0));
//imageBox4.Image = opening;
#endregion
#region gradient
// open > erosion and dilate
//Mat kernel = CvInvoke.GetStructuringElement(ElementShape.Rectangle, new Size(5, 5), new Point(-1, -1));
//var opening = imgColorFiltered.MorphologyEx(MorphOp.Gradient, kernel, new Point(-1, -1), 1, BorderType.Default, new MCvScalar(1.0));
//imageBox4.Image = opening;
#endregion
#region Canny and edge detection
//UMat cannyEdges = new UMat();
//CvInvoke.Canny(imageDilate, cannyEdges, CannyThreshold1, CannyThreshold2, CannyApertureSize, CannyL2Gradient);
//imageBox6.Image = cannyEdges;
#endregion
#region objects
VectorOfVectorOfPoint contours = new VectorOfVectorOfPoint();
CvInvoke.FindContours(opening /*cannyEdges*/, contours, null, RetrType.External, ChainApproxMethod.ChainApproxSimple);
int count = contours.Size;
for (int i = 0; i < count; i++)
{
using (VectorOfPoint contour = contours[i])
using (VectorOfPoint approxContour = new VectorOfPoint())
{
CvInvoke.ApproxPolyDP(contour, approxContour, CvInvoke.ArcLength(contour, true) * 0.05, true);
double area = CvInvoke.ContourArea(approxContour, false);
if (area > 300)
{
CvInvoke.DrawContours(imageFrame, contour, -1, new MCvScalar(255, 0, 0), 2);
imageFrame.Draw($"{area}", new Point(50, 50), Emgu.CV.CvEnum.FontFace.HersheySimplex, 0.8, new Bgr(Color.Red));
}
}
}
#endregion
imageBox1.Image = imageFrame;
//var imgSmoothed = imageFrame.PyrDown().PyrUp();
//imgSmoothed._SmoothGaussian(3);
//imageBox2.Image = imgSmoothed;
//var imgColorFiltered = imgSmoothed.InRange(new Bgr(Bmin, Gmin, Rmin), new Bgr(Bmax, Gmax, Rmax));
//imgColorFiltered = imgColorFiltered.PyrDown().PyrUp();
//imgColorFiltered._SmoothGaussian(3);
//imageBox3.Image = imgColorFiltered;
//Gray grayCannyThreshold = new Gray(GrayCannyTh);
//Gray grayCircleThreshold = new Gray(GrayCircleTh);
//Gray grayLinking = new Gray(80);
//var imgCanny = imgColorFiltered.Canny(grayCannyThreshold.Intensity, grayLinking.Intensity);
//imgColorFiltered.Canny(1.0, 2.0, 3, false);
//var imgCircles = imageFrame.CopyBlank();
//var imgLines = imageFrame.CopyBlank();
//var imgPoly = imageFrame.CopyBlank();
//double dblAccumRes = 1.0;
//double dblMinDist = 1.0;
//int intMinRadius = 50;
//int intMaxRadius = 150;
////var circles = imgColorFiltered.HoughCircles(grayCannyThreshold, grayCircleThreshold, dblAccumRes, dblMinDist, intMinRadius, intMaxRadius)[0];
////foreach (var circ in circles)
////{
//// imgCircles.Draw(circ, new Bgr(Color.Red), 2);
//// imageFrame.Draw(circ, new Bgr(Color.Red), 2);
////}
////Contour<Point> contours = imgCanny.FindContours();
////List<RotatedRect> lstRectangles = new List<RotatedRect>();
//Double dblRhoRes = 1.0;
//Double dblThetaRes = 4.0 * (Math.PI / 180.0);
//int intThreshold = 20;
//Double dblMinLineWidth = 30.0;
//Double dblMinGapBetweenLines = 10.0;
//imgColorFiltered.Ho
//LineSegment2D[] lines = imgCanny.Clone().HoughLinesBinary(dblRhoRes, dblThetaRes, intThreshold, dblMinLineWidth, dblMinGapBetweenLines)[0];
//foreach (LineSegment2D line in lines)
//{
// imgLines.Draw(line, new Bgr(Color.DarkGreen), 2);
// imageFrame.Draw(line, new Bgr(Color.DarkGreen), 2);
//}
//imageBox4.Image = imgLines;
//VectorOfVectorOfPointF contours = new VectorOfVectorOfPointF();
//Mat hierarchy = null;
//CvInvoke.FindContours(imgCanny, contours, hierarchy, Emgu.CV.CvEnum.RetrType.List, Emgu.CV.CvEnum.ChainApproxMethod.ChainCode);
////Contour<Point> contours = imgCanny.FindContours();
//List<MCvBox2D> lstRectangles = new List<MCvBox2D>();
//List<Contour<Point>> lstPoluhons = new List<Contour<Point>>();
//while (contours != null)
//{
// Contour<Point> contour = contours.ApproxPoly(10.0);
// if (contour.Area > 250.0)
// {
// if (contour.Total == 4)
// {
// Point[] ptPoints = contour.ToArray();
// Boolean blnIsRectangle = true;
// LineSegment2D[] ls2dEdges = PointCollection.PolyLine(ptPoints, true);
// for (int i = 0; i < ls2dEdges.Length - 1; i++)
// {
// Double dblAngle = Math.Abs(ls2dEdges[(i + 1) % ls2dEdges.Length].GetExteriorAngleDegree(ls2dEdges[i]));
// if (dblAngle < 80.0 || dblAngle > 100.0)
// {
// blnIsRectangle = false;
// }
// }
// if (blnIsRectangle)
// lstRectangles.Add(contour.GetMinAreaRect());
// }
// }
// contours = contours.HNext;
//}
//foreach (MCvBox2D rect in lstRectangles)
//{
// imgTrisRectsPolys.Draw(rect, new Bgr(Color.Blue), 2);
// if (chbDrawTrianglesAndPolygansOnOriginalImage.Checked == true)
// {
// imgOriginal.Draw(rect, new Bgr(Color.Blue), 2);
// }
//}
//CvInvoke.CvtColor(CameraFeed, HSV, Emgu.CV.CvEnum.ColorConversion.Bgr2Hsv);
//CvInvoke.InRange(HSV, new MCvScalar(Bmin, Gmin, Rmin), new MCvScalar(Bmax, Gmax, Rmax), Threshold);
//CvInvoke.Mor
//histogramBoxCapture. = CameraFeed;
//var imageFrame = Frame.ToImage<Bgr, byte>();
}
catch (Exception ex)
{
}
Busy = false;
}