public void DrawDetectedMarker()
{
// If you want to save markers image, you must change the following values.
const string path = "C:\\detected_markers_6x6_250.png";
using (var image = Image("markers_6x6_250.png", ImreadModes.GrayScale))
using (var outputImage = image.CvtColor(ColorConversionCodes.GRAY2RGB))
using (var dict = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250))
using (var param = DetectorParameters.Create())
{
Point2f[][] corners;
int[] ids;
Point2f[][] rejectedImgPoints;
CvAruco.DetectMarkers(image, dict, out corners, out ids, param, out rejectedImgPoints);
CvAruco.DrawDetectedMarkers(outputImage, corners, ids, new Scalar(255, 0, 0));
CvAruco.DrawDetectedMarkers(outputImage, rejectedImgPoints, null, new Scalar(0, 0, 255));
if (Debugger.IsAttached)
{
Cv2.ImWrite(path, outputImage);
Process.Start(path);
}
}
}
public void Run(cv.Mat src)
{
// The locations of the markers in the image at FilePath.Image.Aruco.
const int upperLeftMarkerId = 160;
const int upperRightMarkerId = 268;
const int lowerRightMarkerId = 176;
const int lowerLeftMarkerId = 168;
var detectorParameters = DetectorParameters.Create();
detectorParameters.CornerRefinementMethod = CornerRefineMethod.Subpix;
detectorParameters.CornerRefinementWinSize = 9;
var dictionary = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict4X4_1000);
CvAruco.DetectMarkers(src, dictionary, out var corners, out var ids, detectorParameters, out var rejectedPoints);
detectedMarkers = src.Clone();
CvAruco.DrawDetectedMarkers(detectedMarkers, corners, ids, cv.Scalar.White);
// Find the index of the four markers in the ids array. We'll use this same index into the
// corners array to find the corners of each marker.
var upperLeftCornerIndex = Array.FindIndex(ids, id => id == upperLeftMarkerId);
var upperRightCornerIndex = Array.FindIndex(ids, id => id == upperRightMarkerId);
var lowerRightCornerIndex = Array.FindIndex(ids, id => id == lowerRightMarkerId);
var lowerLeftCornerIndex = Array.FindIndex(ids, id => id == lowerLeftMarkerId);
// Make sure we found all four markers.
if (upperLeftCornerIndex < 0 || upperRightCornerIndex < 0 || lowerRightCornerIndex < 0 || lowerLeftCornerIndex < 0)
{
return;
}
// Marker corners are stored clockwise beginning with the upper-left corner.
// Get the first (upper-left) corner of the upper-left marker.
var upperLeftPixel = corners[upperLeftCornerIndex][0];
// Get the second (upper-right) corner of the upper-right marker.
var upperRightPixel = corners[upperRightCornerIndex][1];
// Get the third (lower-right) corner of the lower-right marker.
var lowerRightPixel = corners[lowerRightCornerIndex][2];
// Get the fourth (lower-left) corner of the lower-left marker.
var lowerLeftPixel = corners[lowerLeftCornerIndex][3];
// Create coordinates for passing to GetPerspectiveTransform
var sourceCoordinates = new List <cv.Point2f>
{
upperLeftPixel, upperRightPixel, lowerRightPixel, lowerLeftPixel
};
var destinationCoordinates = new List <cv.Point2f>
{
new cv.Point2f(0, 0),
new cv.Point2f(1024, 0),
new cv.Point2f(1024, 1024),
new cv.Point2f(0, 1024),
};
var transform = cv.Cv2.GetPerspectiveTransform(sourceCoordinates, destinationCoordinates);
normalizedImage = new cv.Mat();
cv.Cv2.WarpPerspective(src, normalizedImage, transform, new cv.Size(1024, 1024));
}
void Start () {
// Create default parameres for detection
DetectorParameters detectorParameters = DetectorParameters.Create();
// Dictionary holds set of all available markers
Dictionary dictionary = CvAruco.GetPredefinedDictionary (PredefinedDictionaryName.Dict6X6_250);
// Variables to hold results
Point2f[][] corners;
int[] ids;
Point2f[][] rejectedImgPoints;
// Create Opencv image from unity texture
Mat mat = Unity.TextureToMat (this.texture);
// Convert image to grasyscale
Mat grayMat = new Mat ();
Cv2.CvtColor (mat, grayMat, ColorConversionCodes.BGR2GRAY);
// Detect and draw markers
CvAruco.DetectMarkers (grayMat, dictionary, out corners, out ids, detectorParameters, out rejectedImgPoints);
CvAruco.DrawDetectedMarkers (mat, corners, ids);
// Create Unity output texture with detected markers
Texture2D outputTexture = Unity.MatToTexture (mat);
// Set texture to see the result
RawImage rawImage = gameObject.GetComponent<RawImage> ();
rawImage.texture = outputTexture;
}
public void GetPredefinedDictionary()
{
foreach (PredefinedDictionaryName val in Enum.GetValues(typeof(PredefinedDictionaryName)))
{
var dict = CvAruco.GetPredefinedDictionary(val);
dict.Dispose();
}
}
public void DetectMarkers()
{
using (var image = Image("markers_6x6_250.png", ImreadModes.Grayscale))
using (var dict = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250))
using (var param = DetectorParameters.Create())
{
CvAruco.DetectMarkers(image, dict, out _, out _, param, out _);
}
}
public void GetPredefinedDictionary()
{
#pragma warning disable CS8605
foreach (PredefinedDictionaryName val in Enum.GetValues(typeof(PredefinedDictionaryName)))
#pragma warning restore CS8605
{
var dict = CvAruco.GetPredefinedDictionary(val);
dict.Dispose();
}
}
public void DetectMarkers()
{
using (var image = Image("markers_6x6_250.png", ImreadModes.GrayScale))
using (var dict = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250))
using (var param = DetectorParameters.Create())
{
Point2f[][] corners;
int[] ids;
Point2f[][] rejectedImgPoints;
CvAruco.DetectMarkers(image, dict, out corners, out ids, param, out rejectedImgPoints);
}
}
//private detectObject Detect
void Start()
{
cam = new VideoCapture(source);
cam.Set(CaptureProperty.ConvertRgb, 0); //これが重要!!
cam_frame = new Mat();
leftRight_frame = new Mat[2];
setCamParameters(); //カメラパラメータの設定
ar_dict = CvAruco.GetPredefinedDictionary(dictName);
detect_param = DetectorParameters.Create();
}
public void DictionaryProperties()
{
var dict = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250);
Assert.Equal(250, dict.BytesList.Rows);
Assert.Equal(5, dict.BytesList.Cols); // (6*6 + 7)/8
Assert.Equal(6, dict.MarkerSize);
Assert.Equal(5, dict.MaxCorrectionBits);
dict.MarkerSize = 4;
dict.MaxCorrectionBits = 50;
Assert.Equal(4, dict.MarkerSize);
Assert.Equal(50, dict.MaxCorrectionBits);
}
public static Mat DrawAruco(Mat colorCv)
{
Mat m = new Mat();
colorCv.CopyTo(m);
if (colorCv.Channels() == 4)
{
m = m.CvtColor(ColorConversionCodes.BGRA2BGR);
}
var aruco = FindAruco(m);
CvAruco.DrawDetectedMarkers(m,
aruco.Select(a => a.Points).ToArray(),
aruco.Select(a => a.Id).ToArray());
return(m);
}
// Use this for initialization
void Start()
{
cam = new VideoCapture(source);
cam.Set(CaptureProperty.ConvertRgb, 3);
cam_frame = new Mat();
ar_dict = CvAruco.GetPredefinedDictionary(dictName);
detect_param = DetectorParameters.Create();
cam_Texture = new Texture2D(
cam.FrameWidth, //カメラのwidth
cam.FrameHeight, //カメラのheight
TextureFormat.RGB24, //フォーマットを指定
false //ミニマップ設定
);
this.GetComponent <Renderer>().material.mainTexture = cam_Texture;
}
protected virtual void Init()
{
//For mobile devices which by default have their target frame set to 30
Application.targetFrameRate = 60;
//create OpenCV detector parameters
detectorParameters = DetectorParameters.Create();
detectorParameters.DoCornerRefinement = doCornerRefinement;
//Apply different detection parameters
AssignDetectorParameterStructValues();
dictionary = CvAruco.GetPredefinedDictionary(markerDictionaryType);
timeCount = markerDetectorPauseTime;
//start the marker thread
DetectMarkerAsync();
}
public void DrawMarker()
{
const int markerSidePixels = 128;
const int columns = 4;
const int rows = 5;
const int margin = 20;
// If you want to save markers image, you must change the following values.
const string path = "C:\\markers_6x6_250.png";
int width = columns * markerSidePixels + margin * (columns + 1);
int height = rows * markerSidePixels + margin * (rows + 1);
var id = 0;
Rect roi = new Rect(0, 0, markerSidePixels, markerSidePixels);
using (var outputImage = new Mat(new Size(width, height), MatType.CV_8UC1, Scalar.White))
{
for (var y = 0; y < rows; y++)
{
roi.Top = y * markerSidePixels + margin * (y + 1);
for (var x = 0; x < columns; x++)
{
roi.Left = x * markerSidePixels + margin * (x + 1);
using (var roiMat = new Mat(outputImage, roi))
using (var markerImage = new Mat())
using (var dict = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250))
{
CvAruco.DrawMarker(dict, id++, markerSidePixels, markerImage, 1);
markerImage.CopyTo(roiMat);
}
}
}
if (Debugger.IsAttached)
{
Cv2.ImWrite(path, outputImage);
Process.Start(path);
}
}
}
public void EstimatePoseSingleMarkers()
{
using var image = Image("markers_6x6_250.png", ImreadModes.Grayscale);
using var dict = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250);
var param = DetectorParameters.Create();
CvAruco.DetectMarkers(image, dict, out var corners, out _, param, out _);
using var cameraMatrix = Mat.Eye(3, 3, MatType.CV_64FC1);
using var distCoeffs = Mat.Zeros(4, 1, MatType.CV_64FC1);
using var rvec = new Mat();
using var tvec = new Mat();
using var objPoints = new Mat();
CvAruco.EstimatePoseSingleMarkers(corners, 6, cameraMatrix, distCoeffs, rvec, tvec, objPoints);
Assert.Equal(20, rvec.Total());
Assert.Equal(20, tvec.Total());
Assert.Equal(4, objPoints.Total());
}
private void DetectMarkers()
{
while (true)
{
//On android if there isn't a DebugLog or smth going on here then the thread wont start for some reason.
#if UNITY_ANDROID
Debug.Log("Updating...");
#endif
if (!updateThread)
{
//we skip updating the thread when not needed and also avoids memory exceptions when we disable the
//mono behaviour or we haven't updated the main thread yet!
continue;
}
if (threadCounter > 0)
{
//Gray out the image
Cv2.CvtColor(img, grayedImg, ColorConversionCodes.BGR2GRAY);
//detect the markers and fill the corner and id arrays
CvAruco.DetectMarkers(grayedImg, dictionary, out corners, out ids, detectorParameters,
out rejectedImgPoints);
//throw the callbacks
if (throwMarkerCallbacks)
{
CheckIfLostMarkers();
CheckIfDetectedMarkers();
}
//let the main thread know the process is done
outputImage = true;
Interlocked.Exchange(ref threadCounter, 0);
}
}
}
private bool ProcessTexture(WebCamTexture input, ref Texture2D output,
ARucoUnityHelper.TextureConversionParams textureParameters)
{
imgBuffer = ARucoUnityHelper.TextureToMat(input, textureParameters);
//Debug.Log("New image Assigned");
timeCount += Time.deltaTime;
if (threadCounter == 0 && timeCount >= markerDetectorPauseTime)
{
imgBuffer.CopyTo(img);
Interlocked.Increment(ref threadCounter);
timeCount = 0;
}
updateThread = true;
if (outputImage)
{
if (drawMarkerOutlines)
{
CvAruco.DrawDetectedMarkers(img, corners, ids);
}
output = ARucoUnityHelper.MatToTexture(img, output);
//Debug.Log("Marker image Rendered");
outputImage = false;
}
else
{
output = ARucoUnityHelper.MatToTexture(imgBuffer, output);
//Debug.Log("Camera image Rendered");
}
imgBuffer.Release();
return(true);
}
/// <summary>
/// Finds ARUCO markers in an image and
/// </summary>
/// <param name="colorCv"></param>
/// <returns></returns>
public static List <Marker> FindAruco(Mat colorCv)
{
Mat m = new Mat();
colorCv.CopyTo(m);
if (colorCv.Channels() == 4)
{
m = m.CvtColor(ColorConversionCodes.BGRA2BGR);
}
//Kinect color image is flipped, correct before ARUCO detection
var flipped = m.Flip(FlipMode.Y);
List <Marker> markers = new List <Marker>();
var dictionary = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250);
Point2f[][] points;
Point2f[][] rejPoints;
int[] ids;
var parms = DetectorParameters.Create();
CvAruco.DetectMarkers(flipped, dictionary, out points, out ids, parms, out rejPoints);
for (int i = 0; i < ids.Length; i++)
{
var id = ids[i];
var arucoMarkerCorners = points[i];
for (int j = 0; j < arucoMarkerCorners.Length; j++)
{
arucoMarkerCorners[j].X = KinectSettings.COLOR_WIDTH - arucoMarkerCorners[j].X - 1;
}
var area = Cv2.ContourArea(arucoMarkerCorners);
markers.Add(new Marker()
{
Id = id, Points = arucoMarkerCorners, Area = area
});
}
return(markers.OrderByDescending(ma => ma.Area).ToList());
}
static void Main(string[] args)
{
//取得した画像とその差分画像を表示する
#region
//Mat src = new Mat(@"C:\Users\tyani\Downloads\lenna.png", ImreadModes.Grayscale);
//// Mat src = Cv2.ImRead("lenna.png", ImreadModes.Grayscale);
//Mat dst = new Mat();
//Cv2.Canny(src, dst, 50, 200);
//using (new Window("src image", src))
//using (new Window("dst image", dst))
//{
// Cv2.WaitKey();
//}
#endregion
//キャプチャーした映像から顔を認識してその位置に丸を付ける
#region
//using (var win = new Window("capture"))
//using (var video = new VideoCapture(0))
//{
// //保存先行列
// var frame = new Mat();
// //分類器
// var haarcascade = new CascadeClassifier(@"C:\Users\tyani\Downloads\opencv-master\data\haarcascades\haarcascade_frontalface_default.xml");
// //グレースケール
// var gray = new Mat();
// while (true)
// {
// video.Read(frame);
// //grayにframeのグレースケール返還画像を代入
// Cv2.CvtColor(frame, gray, ColorConversionCodes.BGR2GRAY);
// //検出
// var faces = haarcascade.DetectMultiScale(gray);
// //検出箇所に円を描画
// foreach (OpenCvSharp.Rect face in faces)
// {
// var center = new Point
// {
// X = (int)(face.X + face.Width * 0.5),
// Y = (int)(face.Y + face.Height * 0.5)
// };
// var axes = new Size
// {
// Width = (int)(face.Width * 0.5),
// Height = (int)(face.Height * 0.5)
// };
// Cv2.Ellipse(frame, center, axes, 0, 0, 360, Scalar.Red, 4);
// Console.WriteLine("Found!");
// }
// win.ShowImage(frame);
// if (Cv2.WaitKey(30) >= 0) { break; }
// }
//}
#endregion
//カメラキャリブレーションを行う
#region
////チェス盤のコーナー座標がを読み込めた数
//int imgInd = 0;
////読み込むチェス盤の定義
//const int BOARD_W = 7;
//const int BOARD_H = 7;
//Size BOARD_SIZE = new Size(BOARD_W, BOARD_H);
//const int SCALE = 26;
////精度の低いコーナーが検出できた座標のリスト
//var imageCorners = new Mat<Point2f>();
////精度の高いコーナー座標の配列
//Point2f[] imageCorners2;
////検出できたframe中のコーナー座標のリスト
//var imagePoints = new List<Mat<Point2f>>();
////毎秒のフレーム画像
//var frame = new Mat();
////フレーム画像のグレースケール
//var gray = new Mat();
////認識したチェス盤画像
//var chessboard = new List<Mat>();
////精度の高いコーナー座標を作る際の終了基準
//var criteria = new TermCriteria(CriteriaType.Count | CriteriaType.Eps, 30, 0.001);
////ウィンドウを生成
//using (var capWindow = new Window("capture"))
//using (var chessWindow = new Window("DrawChessboardCorner"))
////videoにカメラから取得した映像を代入
//using (var video = new VideoCapture(0))
//{
// while (true)
// {
// video.Read(frame);
// var key = Cv2.WaitKey(1);
// if (key == 'c')
// {
// Cv2.CvtColor(frame, gray, ColorConversionCodes.BGR2GRAY);
// //チェス盤のコーナーが検出できたらretにtrueを代入
// var ret = Cv2.FindChessboardCorners(frame, BOARD_SIZE, imageCorners);
// //Console.WriteLine(imageCorners.Get<Point2f>(0));
// if (ret)
// {
// Console.WriteLine("find chess board corners.");
// chessboard.Add(frame);
// //精度を高めたコーナー座標をimageCorners2に代入
// imageCorners2 = Cv2.CornerSubPix(gray, imageCorners, new Size(9, 9), new Size(-1, -1), criteria);
// //Console.WriteLine(imageCorners2[0]);
// Cv2.DrawChessboardCorners(frame, BOARD_SIZE, imageCorners2, ret);
// imageCorners.SetArray<Point2f>(imageCorners2);
// //Console.WriteLine(imageCorners.Get<Point2f>(0));
// //Console.WriteLine(imageCorners.Size());
// imagePoints.Add(imageCorners);
// chessWindow.ShowImage(frame);
// Cv2.WaitKey(500);
// imgInd++;
// }
// }
// else if (key == 'q')
// {
// break;
// }
// else
// {
// Cv2.PutText(frame, "Number of caputure: " + imgInd.ToString(), new Point(30, 20), HersheyFonts.HersheyPlain, 1, new Scalar(0, 255, 0));
// Cv2.PutText(frame, "c: Capture the image", new Point(30, 40), HersheyFonts.HersheyPlain, 1, new Scalar(0, 255, 0));
// Cv2.PutText(frame, "q: Finish capturing and calcurate the camera matrix and distortion", new Point(30, 60), HersheyFonts.HersheyPlain, 1, new Scalar(0, 255, 0));
// capWindow.ShowImage(frame);
// }
// }
// //よくわからない
// var objectPoints = new List<Mat<Point3f>>();
// var objectConrers = new Mat<Point3f>();
// for (int j = 0; j < BOARD_H; j++)
// {
// for (int i = 0; i < BOARD_W; i++)
// {
// objectConrers.PushBack(new Point3f(i * SCALE, j * SCALE, 0f));
// }
// }
// for (int i = 0; i < imgInd; i++)
// objectPoints.Add(objectConrers);
// var cameraMatrix = new Mat();
// var distCoeffs = new Mat();
// Mat[] rvecs;
// Mat[] tvecs;
// var rms = Cv2.CalibrateCamera(objectPoints, imagePoints, frame.Size(), cameraMatrix, distCoeffs, out rvecs, out tvecs);
// Console.WriteLine("Re-projection Error(unit: pixel) : " + rms);
// Console.WriteLine("cameraMatrix(unit: pixel) : " + cameraMatrix);
// Console.WriteLine("distCoeffs : " + distCoeffs);
// var dst = new Mat();
// Cv2.Undistort(chessboard[0], dst, cameraMatrix, distCoeffs);
// using (new Window("補正画像", WindowMode.AutoSize, dst))
// {
// while (true)
// if (Cv2.WaitKey() == 'q')
// break;
// }
//}
#endregion
//SIFTマッチング
#region
//var leftFrame = new Mat();
//var rightFrame = new Mat();
//var view = new Mat();
//var leftGray = new Mat();
//var rightGray = new Mat();
//var sift = SIFT.Create();
//KeyPoint[] keypoints1, keypoints2;
//var descriptors1 = new Mat();
//var descriptors2 = new Mat();
//using (var leftCapWindow = new Window("left capture"))
//using (var rightCapWindow = new Window("right capture"))
//using (var leftVideo = new VideoCapture(0))
//using (var rightVideo = new VideoCapture(1))
//{
// while (true)
// {
// leftVideo.Read(leftFrame);
// rightVideo.Read(rightFrame);
// Cv2.PutText(leftFrame, "d: detect and compute SIFT", new Point(30, 40), HersheyFonts.HersheyPlain, 1, new Scalar(0, 255, 0));
// Cv2.PutText(rightFrame, "d: detect and compute SIFT", new Point(30, 40), HersheyFonts.HersheyPlain, 1, new Scalar(0, 255, 0));
// leftCapWindow.ShowImage(leftFrame);
// rightCapWindow.ShowImage(rightFrame);
// var key = Cv2.WaitKey(1);
// if (key == 'd')
// {
// Cv2.CvtColor(leftFrame, leftGray, ColorConversionCodes.BGR2GRAY);
// Cv2.CvtColor(rightFrame, rightGray, ColorConversionCodes.BGR2GRAY);
// sift.DetectAndCompute(leftGray, null, out keypoints1, descriptors1);
// sift.DetectAndCompute(rightGray, null, out keypoints2, descriptors2);
// BFMatcher matcher = new BFMatcher(NormTypes.L2, false);
// DMatch[] matches = matcher.Match(descriptors1, descriptors2);
// Cv2.DrawMatches(leftGray, keypoints1, rightGray, keypoints2, matches, view);
// using (new Window("SIFT matching", WindowMode.AutoSize, view))
// {
// Cv2.WaitKey();
// }
// }
// }
//}
#endregion
//視差マップ生成
#region
//var leftFrame = new Mat();
//var rightFrame = new Mat();
//var disparity = new Mat();
//var leftGray = new Mat();
//var rightGray = new Mat();
//var sift = SIFT.Create();
//KeyPoint[] keypoints1, keypoints2;
//var descriptors1 = new Mat();
//var descriptors2 = new Mat();
//using (var leftCapWindow = new Window("left capture"))
//using (var rightCapWindow = new Window("right capture"))
//using (var leftVideo = new VideoCapture(1))
//using (var rightVideo = new VideoCapture(0))
//{
// while (true)
// {
// leftVideo.Read(leftFrame);
// rightVideo.Read(rightFrame);
// leftCapWindow.ShowImage(leftFrame);
// rightCapWindow.ShowImage(rightFrame);
// Cv2.CvtColor(leftFrame, leftGray, ColorConversionCodes.BGR2GRAY);
// Cv2.CvtColor(rightFrame, rightGray, ColorConversionCodes.BGR2GRAY);
// var stereo = StereoSGBM.Create(0, 16*10, 11, mode:StereoSGBMMode.HH);
// stereo.Compute(leftGray, rightGray, disparity);
// disparity.Normalize(alpha: 0, beta: 255, normType: NormTypes.MinMax, dtype: MatType.CV_8U);
// using (new Window("disparity map", WindowMode.AutoSize, disparity))
// {
// var key = Cv2.WaitKey(1);
// if (key == 'q')
// break;
// }
// }
//}
#endregion
//もっとも簡単なStereoMatchingのTutorial
#region
//var imgL = new Mat(@"C:\Users\tyani\Downloads\scene1.row3.col1.png", ImreadModes.Grayscale);
//var imgR = new Mat(@"C:\Users\tyani\Downloads\scene1.row3.col3.png", ImreadModes.Grayscale);
//var disparity = new Mat();
//var stereo = StereoBM.Create();
//stereo.Compute(imgL, imgR, disparity);
//using(new Window(disparity))
//{
// Cv2.WaitKey(0);
//}
#endregion
//ARマーカー
#region
using (var video0 = new VideoCapture(0))
using (var video1 = new VideoCapture(1))
using (var window0 = new Window("capture0"))
using (var window1 = new Window("capture1"))
{
var dictionary = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250);
var parameters = DetectorParameters.Create();
var frames = new List <Mat> {
new Mat(), new Mat()
};
var videos = new List <VideoCapture>();
var windows = new List <Window>();
Point2f[][] corners;
int[] ids;
int[] previousIds;
Point2f[][] rejectedImgPoints;
videos.Add(video0);
videos.Add(video1);
windows.Add(window0);
windows.Add(window1);
var wasFoundList = new List <bool> {
false, false
};
var isTouchedList = new List <bool> {
false, false
};
var wasTouched = false;
while (true)
{
for (int i = 0; i < 2; i++)
{
videos[i].Read(frames[i]);
CvAruco.DetectMarkers(frames[i], dictionary, out corners, out ids, parameters, out rejectedImgPoints);
isTouchedList[i] = wasFoundList[i] && !(ids.Length > 0);
if (ids.Length > 0)
{
wasFoundList[i] = true;
CvAruco.DrawDetectedMarkers(frames[i], corners, ids);
}
else
{
wasFoundList[i] = false;
isTouchedList[i] = true;
}
windows[i].ShowImage(frames[i]);
}
if (!isTouchedList.Contains(false))
{
if (!wasTouched)
{
Console.WriteLine("Hello world!");
for (int i = 0; i < isTouchedList.Count; i++)
{
isTouchedList[i] = false;
}
}
wasTouched = true;
}
else
{
wasTouched = false;
}
var key = Cv2.WaitKey(1);
if (key == 'q')
{
break;
}
}
}
#endregion
#region
//var numCamera = 1;
//var dictionary = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict7X7_50);
//var parameters = DetectorParameters.Create();
//var videoCaps = new List<VideoCapture>(numCamera);
//var windows = new List<Window>(numCamera);
//var capFrames = new List<Mat>(numCamera);
////var markerCornersList = new List<Point2f[][]>(numCamera);
////var markerIdsList = new List<int[]>(numCamera);
////var rejectedImgPointsList = new List<Point2f[][]>(numCamera);
//var isFoundList = new List<bool>(numCamera);
//var isTouchedList = new List<bool>(numCamera);
//try
//{
// for (int i = 0; i < numCamera; i++)
// {
// videoCaps.Add(new VideoCapture(i));
// windows.Add(new Window("VideoCapture" + i.ToString()));
// }
// while (true)
// {
// for (int i = 0; i < numCamera; i++)
// {
// var capFrame = new Mat();
// Point2f[][] markerCorners;
// int[] markerIds;
// Point2f[][] rejectedImgPoints;
// videoCaps[i].Read(capFrame);
// CvAruco.DetectMarkers(capFrame, dictionary, out markerCorners, out markerIds, parameters, out rejectedImgPoints);
// if (!isFoundList[i] && markerIds.Length > 0)
// {
// isFoundList[i] = true;
// }
// if (isFoundList[i] && !(markerIds[i] > 0))
// {
// isTouchedList[i] = true;
// }
// if (markerIds.Length > 0)
// {
// CvAruco.DrawDetectedMarkers(capFrame, markerCorners, markerIds);
// }
// windows[i].ShowImage(capFrame);
// }
// if (!isTouchedList.Contains(false))
// {
// Console.WriteLine("Hello world!");
// }
// var key = Cv2.WaitKey(1);
// if (key == 'q')
// {
// break;
// }
// }
//}
//catch
//{
// Console.WriteLine("例外：カメラが取得できませんでした。");
//}
//finally
//{
// for (int i = 0; i < numCamera; i++)
// {
// if (videoCaps[i] != null)
// videoCaps[i].Dispose();
// if (windows[i] != null)
// {
// windows[i].Dispose();
// }
// }
//}
//test
#endregion
}
// Update is called once per frame
void Update()
{
Point2f[][] maker_corners; //ARマーカのカドの座標
int[] maker_ids; //検出されたARマーカのID
Point2f[][] reject_points;
cam.Read(cam_frame);//フレームの更新
//ARマーカの検出
CvAruco.DetectMarkers(cam_frame, ar_dict, out maker_corners, out maker_ids, detect_param, out reject_points);
//キャリブレーションデータを設定
double[,] array3_3 = new double[3, 3] {
{ 548.8, 0.0, 332.1 }, { 0.0, 549.8, 236.2 }, { 0.0, 0.0, 1.0 }
};
Mat cameraMatrix = new Mat(3, 3, MatType.CV_32FC1, array3_3);
double[,] array1_5 = new double[1, 5] {
{ -0.0886, 0.4802, -0.0026, 0.0019, -0.8238 }
};
Mat distortionCoefficients = new Mat(1, 5, MatType.CV_32FC1, array1_5);
Mat rvec = new Mat();
Mat tvec = new Mat();
//VectorOfVec3f vector3_rvec = new VectorOfVec3f();
//IEnumerable<VectorOfVec3f> vector3_rvec = new IEnumerable<VectorOfVec3f>;
if (maker_ids.Length != 0)
{
//検出されたマーカ情報の描画
CvAruco.DrawDetectedMarkers(cam_frame, maker_corners, maker_ids, new Scalar(0, 255, 0));
//Debug.Log("Find");
/*
* Mat rvec = new Mat(3, maker_ids.Length, MatType.CV_32FC1);
* Mat tvec = new Mat(3, maker_ids.Length, MatType.CV_32FC1);
*/
/*
* Debug.Log(tvec.Rows);
* Debug.Log(tvec.Cols);
* Debug.Log(tvec.Size());
* Debug.Log(tvec.Dims());
*/
CvAruco.EstimatePoseSingleMarkers(maker_corners, (float)1.8, cameraMatrix, distortionCoefficients, rvec, tvec);
//マーカまでの距離
/*
* //1
* Debug.Log(tvec.Rows);
* //1
* Debug.Log(tvec.Cols);
* //1:1
* Debug.Log(tvec.Size());
* //2
* Debug.Log(tvec.Dims());
* //3=#define CV_16S
* Debug.Log(tvec.Channels());
*/
Vec3d vec3D = tvec.At <Vec3d>(0, 0);
double z = vec3D[2];
Debug.Log(z);
/*
* Debug.Log(tvec.At<Vec3d>(0, 0)[2]);
*/
//Debug.Log(tvec.Data[0]);
//Debug.Log(tvec.At<Vec3b>(1, 1)[2]);
//Debug.Log(tvec[new OpenCvSharp.Rect(0, 1,0,1)]);
/*
* Debug.Log(rvec.Rows);
* Debug.Log(rvec.Cols);
* Debug.Log(rvec.Size());
* Debug.Log(rvec.Dims());
*/
/*
* //Debug.Log(cameraMatrix.Size());
* //1
* Debug.Log(distortionCoefficients.Rows);
* //5
* Debug.Log(distortionCoefficients.Cols);
* Debug.Log(distortionCoefficients.Size());
*/
/*
* //1channelはdoubleできれいに取れる
* Debug.Log(distortionCoefficients.At<double>(0,0));
* //1
* Debug.Log(distortionCoefficients.Channels());
*/
//Debug.Log(tvec.)
//24
//Debug.Log(tvec.Step());
//3
//Debug.Log(tvec.Channels());
//Debug.Log(tvec.At<double>(0, 0));
//Debug.Log(tvec.At<double>(0, 1));
//Debug.Log(tvec.At<double>(0, 2));
/*
* Debug.Log(rvec.GetArray(0, 1));
* Debug.Log(rvec.GetArray(0, 2));
*/
}
cam_Texture.LoadImage(cam_frame.ImEncode());
}
void Update()
{
//Leapmotionのカメラの映像を取得
readLeftRightFrame();
Point2f[][] maker_corners;
int[] maker_ids;
Point2f[][] reject_points;
//ARマーカの検出
CvAruco.DetectMarkers(leftRight_frame[0], ar_dict, out maker_corners, out maker_ids, detect_param, out reject_points);
GameObject[] animals = GameObject.FindGameObjectsWithTag("animal");
//マーカが検出されない場合はすべてのanimalオブジェクトを削除
if (maker_ids.Length == 0)
{
foreach (GameObject animal in animals)
{
animalMaster animal_script = animal.GetComponent <animalMaster>();
animal_script.destoryThis();
Debug.Log("destroy");
}
}
//マーカが検出されなかったanimalオブジェクトを削除
else
{
foreach (GameObject animal in animals)
{
//検出されたIDと一致するゲームオブジェクトが存在しなかったら
if (DetObj.CheckSameID(animal, maker_ids) == false)
{
animalMaster animal_script = animal.GetComponent <animalMaster>();
animal_script.destoryThis();
}
}
}
if (maker_ids.Length > 0)
{
//検出されたマーカ情報の描画
CvAruco.DrawDetectedMarkers(leftRight_frame[0], maker_corners, maker_ids, new Scalar(0, 255, 0));
//マーカの姿勢推定
OutputArray tvec = new Mat(), rvec = new Mat();
CvAruco.EstimatePoseSingleMarkers(maker_corners, 0.6f, cameraMatrix, distCoefficients, rvec, tvec, null);
//検出されたIDに対する操作
for (int i = 0; i < maker_ids.Length; i++)
{
//ID辞書と一致するマーカが検出されたら
if (DetObj.CheckTargetMarker(maker_ids[i], id_dict) == true)
{
if (animals != null)
{
//すでにオブジェクトが存在していたら
if (DetObj.CheckTargetObject(maker_ids[i], animals) == true)
{
int animal_index = DetObj.GetAnimalIndex(maker_ids[i], animals);
animalMaster animal_script = animals[animal_index].GetComponent <animalMaster>();
animal_script.moveModel(i, tvec, rvec);
}
//オブジェクトが存在しない場合
else
{
GameObject newAnimal = Instantiate(animal);
animalMaster animal_script = newAnimal.GetComponent <animalMaster>();
int animal_index = DetObj.GetAnimalIndex(maker_ids[i], animals);
animal_script.setParam(maker_ids[i], i, tvec, rvec);
}
}
else
{
GameObject newAnimal = Instantiate(animal);
animalMaster animal_script = newAnimal.GetComponent <animalMaster>();
animal_script.setParam(maker_ids[i], 0, tvec, rvec);
}
}
}
}
}
/// <summary>
/// Detect markers.
/// </summary>
/// <param name="pixels">
/// The image where to detect markers.
/// For example, you can use get pixels from web camera https://docs.unity3d.com/ScriptReference/WebCamTexture.GetPixels32.html
/// </param>
public List <int> Detect(Mat mat, int width, int height)
{
List <int> result = new List <int> ();
markerTransforms.Clear();
// Create default parameres for detection
DetectorParameters detectorParameters = DetectorParameters.Create();
// Dictionary holds set of all available markers
Dictionary dictionary = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250);
// Variables to hold results
Point2f[][] corners;
int[] ids;
Point2f[][] rejectedImgPoints;
// Convert image to grasyscale
Mat grayMat = new Mat();
Cv2.CvtColor(mat, grayMat, ColorConversionCodes.BGR2GRAY);
// Detect markers
CvAruco.DetectMarkers(grayMat, dictionary, out corners, out ids, detectorParameters, out rejectedImgPoints);
// CvAruco.DrawDetectedMarkers (mat, corners, ids);
float markerSizeInMeters = 1f;
Point3f[] markerPoints = new Point3f[] {
new Point3f(-markerSizeInMeters / 2f, markerSizeInMeters / 2f, 0f),
new Point3f(markerSizeInMeters / 2f, markerSizeInMeters / 2f, 0f),
new Point3f(markerSizeInMeters / 2f, -markerSizeInMeters / 2f, 0f),
new Point3f(-markerSizeInMeters / 2f, -markerSizeInMeters / 2f, 0f)
};
double max_wh = (double)Math.Max(width, height);
double fx = max_wh;
double fy = max_wh;
double cx = width / 2.0d;
double cy = height / 2.0d;
//
double[,] cameraMatrix = new double[3, 3] {
{ fx, 0d, cx },
{ 0d, fy, cy },
{ 0d, 0d, 1d }
};
double[] distCoeffs = new double[4] {
0d, 0d, 0d, 0d
};
double[] rvec = new double[3] {
0d, 0d, 0d
};
double[] tvec = new double[3] {
0d, 0d, 0d
};
double[,] rotMat = new double[3, 3] {
{ 0d, 0d, 0d }, { 0d, 0d, 0d }, { 0d, 0d, 0d }
};
for (int i = 0; i < ids.Length; i++)
{
Cv2.SolvePnP(markerPoints, corners[i], cameraMatrix, distCoeffs, out rvec, out tvec, false, SolvePnPFlags.Iterative);
// CvAruco.DrawAxis(mat, cameraMatrix, distCoeffs, rvec, tvec, 1.0f);
Cv2.Rodrigues(rvec, out rotMat);
Matrix4x4 matrix = new Matrix4x4();
matrix.SetRow(0, new Vector4((float)rotMat[0, 0], (float)rotMat[0, 1], (float)rotMat[0, 2], (float)tvec[0]));
matrix.SetRow(1, new Vector4((float)rotMat[1, 0], (float)rotMat[1, 1], (float)rotMat[1, 2], (float)tvec[1]));
matrix.SetRow(2, new Vector4((float)rotMat[2, 0], (float)rotMat[2, 1], (float)rotMat[2, 2], (float)tvec[2]));
matrix.SetRow(3, new Vector4(0f, 0f, 0f, 1f));
result.Add(ids[i]);
markerTransforms.Add(matrix);
}
return(result);
}
public static void BuildRegistrationCube(string path, double cmWidth = 10.4)
{
var definition = CoordinateDefinition.Cube(11 * 2.5, cmWidth);
var dir = Path.GetDirectoryName(path);
var sides = new Dictionary <string, int>() //side, starting index id
{
{ "Side1", 0 },
{ "Side2", 4 },
{ "Side3", 8 },
{ "Side4", 12 },
{ "Side5", 16 },
{ "Side6", 20 },
};
var pixelsPerCM = 300 / 2.54;
//2 SQUARES WIDE, 2 SQUARES HIGH, 150 margin on width,
//100 margin on height
var paperPxWidth = pixelsPerCM * cmWidth; //11in - 1in margin *300px/in
var paperPxHeight = pixelsPerCM * cmWidth; //17in - 1in margin *300px/in
var markerPxWidth = (int)(pixelsPerCM * 4.5 / 10.4 * cmWidth);
var marginWidth = (int)((paperPxWidth - 2.0 * markerPxWidth) / 4);
var marginHeight = marginWidth;
var rows = 2;
var columns = 2;
var id = 0;
foreach (var side in sides)
{
Rect roi = new Rect(0, 0, markerPxWidth, markerPxWidth);
using (var outputImage = new Mat(new Size(paperPxWidth, paperPxHeight), MatType.CV_8UC1, Scalar.White))
{
for (var y = 0; y < rows; y++)
{
roi.Top = y * markerPxWidth + marginHeight * (y * 2 + 1);
for (var x = 0; x < columns; x++)
{
roi.Left = x * markerPxWidth + marginWidth * (x * 2 + 1);
using (var roiMat = new Mat(outputImage, roi))
using (var markerImage = new Mat())
using (var dict2 = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250))
{
CvAruco.DrawMarker(dict2, id++, markerPxWidth, markerImage, 1);
markerImage.CopyTo(roiMat);
}
}
}
var crossHairWidth = 1 * pixelsPerCM; // 1 cm
var crossHairColor = new Scalar(25, 25, 25);
Cv2.Line(outputImage, new Point((float)paperPxWidth / 2, (float)paperPxHeight / 2), new Point((float)paperPxWidth / 2, (float)(paperPxHeight / 2 - crossHairWidth)), crossHairColor);
Cv2.Line(outputImage, new Point((float)paperPxWidth / 2, (float)paperPxHeight / 2), new Point((float)paperPxWidth / 2, (float)(paperPxHeight / 2 + crossHairWidth)), crossHairColor);
Cv2.Line(outputImage, new Point((float)paperPxWidth / 2, (float)paperPxHeight / 2), new Point((float)(paperPxWidth / 2 - crossHairWidth), (float)(paperPxHeight / 2)), crossHairColor);
Cv2.Line(outputImage, new Point((float)paperPxWidth / 2, (float)paperPxHeight / 2), new Point((float)(paperPxWidth / 2 + crossHairWidth), (float)(paperPxHeight / 2)), crossHairColor);
Cv2.PutText(outputImage, side.Key, new Point((float)paperPxWidth / 2 - marginWidth / 1.3, (float)paperPxHeight), HersheyFonts.HersheyPlain, 1.5 / 10.4 * cmWidth, new Scalar(25, 25, 25), 1);
path = Path.Combine(dir, side.Key + ".png");
Cv2.ImWrite(path, outputImage);
}
}
}
public void DetectTouchByTwoCamera(Action action)
{
using (var video0 = new VideoCapture(0))
using (var video1 = new VideoCapture(1))
using (var window0 = new Window("capture0"))
using (var window1 = new Window("capture1"))
{
var dictionary = CvAruco.GetPredefinedDictionary(PredefinedDictionaryName.Dict6X6_250);
var parameters = DetectorParameters.Create();
var frames = new List <Mat> {
new Mat(), new Mat()
};
var videos = new List <VideoCapture>();
var windows = new List <Window>();
Point2f[][] corners;
int[] ids;
int[] previousIds;
Point2f[][] rejectedImgPoints;
videos.Add(video0);
videos.Add(video1);
windows.Add(window0);
windows.Add(window1);
var wasFoundList = new List <bool> {
false, false
};
var isTouchedList = new List <bool> {
false, false
};
var wasTouched = false;
while (true)
{
for (int i = 0; i < 2; i++)
{
videos[i].Read(frames[i]);
CvAruco.DetectMarkers(frames[i], dictionary, out corners, out ids, parameters, out rejectedImgPoints);
isTouchedList[i] = wasFoundList[i] && !(ids.Length > 0);
if (ids.Length > 0)
{
wasFoundList[i] = true;
CvAruco.DrawDetectedMarkers(frames[i], corners, ids);
}
else
{
wasFoundList[i] = false;
isTouchedList[i] = true;
}
windows[i].ShowImage(frames[i]);
}
if (!isTouchedList.Contains(false))
{
if (!wasTouched)
{
Console.WriteLine("Hello world!");
for (int i = 0; i < isTouchedList.Count; i++)
{
isTouchedList[i] = false;
}
}
wasTouched = true;
}
else
{
wasTouched = false;
}
var key = Cv2.WaitKey(1);
if (key == 'q')
{
break;
}
}
}
}
