public static Rectangle GetRect(VectorOfKeyPoint vKeyPoint)
{
//Rectangle ret = new Rectangle(0,0,0,0);
int minx = 10000;
int miny = 10000;
int maxx = 0;
int maxy = 0;
for (int i = 0; i < vKeyPoint.Size; i++)
{
MKeyPoint p = vKeyPoint[i];
if (p.Point.X < minx)
{
minx = (int)p.Point.X;
}
if (p.Point.X > maxx)
{
maxx = (int)p.Point.X;
}
if (p.Point.Y < miny)
{
miny = (int)p.Point.Y;
}
if (p.Point.Y > maxy)
{
maxy = (int)p.Point.Y;
}
}
return(new Rectangle(minx, miny, maxx - minx, maxy - miny));
}
/// <summary>
/// Convert the standard vector to an array of float
/// </summary>
/// <returns>An array of float</returns>
public MKeyPoint[] ToArray()
{
MKeyPoint[] res = new MKeyPoint[Size];
GCHandle handle = GCHandle.Alloc(res, GCHandleType.Pinned);
VectorOfKeyPointCopyData(_ptr, handle.AddrOfPinnedObject());
handle.Free();
return(res);
}
/// <summary>
/// Get the item in the specific index
/// </summary>
/// <param name="index">The index</param>
/// <returns>The item in the specific index</returns>
public MKeyPoint this[int index]
{
get
{
MKeyPoint result = new MKeyPoint();
VectorOfKeyPointGetItem(_ptr, index, ref result);
return(result);
}
}
protected override VideoCloudPoints ReadFrom(string path)
{
using (var file = File.OpenRead(path))
{
using (var stream = new BinaryReader(file))
{
var videoPath = stream.ReadString();
var frameCounts = stream.ReadInt32();
var videCloudPoints = new VideoCloudPoints(videoPath, frameCounts);
for (int i = 0; i < frameCounts; i++)
{
var keysCount = stream.ReadInt32();
var array = new MKeyPoint[keysCount];
for (int j = 0; j < keysCount; j++)
{
var keyFeature = new MKeyPoint();
// keyFeature.Angle = stream.ReadSingle();
// keyFeature.ClassId = stream.ReadInt32();
// keyFeature.Octave = stream.ReadInt32();
// keyFeature.Response = stream.ReadSingle();
// keyFeature.Size = stream.ReadSingle();
keyFeature.Point = new PointF(stream.ReadSingle(), stream.ReadSingle());
array[j] = keyFeature;
}
videCloudPoints.SetKeyFeatures(i, new VectorOfKeyPoint(array));
}
int matchesListCount = stream.ReadInt32();
for (int i = 0; i < matchesListCount; i++)
{
int matchesCount = stream.ReadInt32();
int firstFrameIndex = -1;
int firstFeatureIndex = -1;
for (int j = 0; j < matchesCount; j++)
{
int frameIndex = stream.ReadInt32();
int featureIndex = stream.ReadInt32();
if (firstFeatureIndex != -1 && firstFeatureIndex != -1)
{
videCloudPoints.Unite(firstFrameIndex, firstFeatureIndex, frameIndex, featureIndex);
}
else
{
firstFrameIndex = frameIndex;
firstFeatureIndex = featureIndex;
}
}
}
return(videCloudPoints);
}
}
}
/// <summary>
/// Convert the standard vector to an array of KeyPoint
/// </summary>
/// <returns>An array of KeyPoint</returns>
public MKeyPoint[] ToArray()
{
MKeyPoint[] res = new MKeyPoint[Size];
if (res.Length > 0)
{
GCHandle handle = GCHandle.Alloc(res, GCHandleType.Pinned);
CvInvoke.VectorOfKeyPointCopyData(_ptr, handle.AddrOfPinnedObject());
handle.Free();
}
return(res);
}
public static VectorOfKeyPoint GetKeyPointsVector(VectorOfPointF points)
{
var keyPoints = new MKeyPoint[points.Size];
for (int i = 0; i < points.Size; i++)
{
keyPoints[i] = new MKeyPoint()
{
Point = points[i],
Size = 1,
Angle = 0,
Response = 0
};
}
return(new VectorOfKeyPoint(keyPoints));
}
static object[] DetectAndCompute2(UMat img, Feature2D detector, Feature2D computer) // находит и обрабатывает дескрипторы изображения
{
object[] outp = new object[0];
UMat descriptors = new UMat();
var mkp = new MKeyPoint[0];
VectorOfKeyPoint keypoints;
try
{
mkp = detector.Detect(img);
keypoints = new VectorOfKeyPoint(mkp);
computer.Compute(img, keypoints, descriptors);
outp = new object[] { keypoints, descriptors };
}
finally
{
}
return(outp);
}
public Mat FingerprintDescriptor(Mat input)
{
var harris_normalised = PrepareImage(input);
float threshold = 125.0f;
List <MKeyPoint> mKeyPoints = new List <MKeyPoint>();
Mat rescaled = new Mat();
VectorOfKeyPoint keypoints = new VectorOfKeyPoint();
double scale = 1.0, shift = 0.0;
CvInvoke.ConvertScaleAbs(harris_normalised, rescaled, scale, shift);
Mat[] mat = new Mat[] { rescaled, rescaled, rescaled };
VectorOfMat vectorOfMat = new VectorOfMat(mat);
int[] from_to = { 0, 0, 1, 1, 2, 2 };
Mat harris_c = new Mat(rescaled.Size, DepthType.Cv8U, 3);
CvInvoke.MixChannels(vectorOfMat, harris_c, from_to);
for (int x = 0; x < harris_c.Width; x++)
{
for (int y = 0; y < harris_c.Height; y++)
{
if (GetFloatValue(harris_c, x, y) > threshold)
{
MKeyPoint m = new MKeyPoint
{
Size = 1,
Point = new PointF(x, y)
};
mKeyPoints.Add(m);
}
}
}
keypoints.Push(mKeyPoints.ToArray());
Mat descriptors = new Mat();
ORBDetector ORBCPU = new ORBDetector();
ORBCPU.Compute(_input_thinned, keypoints, descriptors);
return(descriptors);
}
static void test_match()
{
Bitmap b1 = new Bitmap(@"C:\test\test_1\temp_menu.jpg");
Bitmap sl = new Bitmap(@"C:\test\scroll_left.jpg");
Image <Gray, Byte> slicon = new Image <Gray, byte>(sl);
slicon = slicon.Not();
slicon.Save("temp_1.jpg");
Image <Gray, Byte> test = new Image <Gray, Byte>(b1);
//long l;
//Mat r = DrawMatches.Draw(slicon.Mat, test.Mat, out l);
//r.Save("temp_2.jpg");
//SURF surfCPU = new SURF(400);
//Brisk surfCPU = new Brisk();
SIFT surfCPU = new SIFT();
VectorOfKeyPoint modelKeyPoints = new VectorOfKeyPoint();
VectorOfKeyPoint observedKeyPoints = new VectorOfKeyPoint();
UMat modelDescriptors = new UMat();
UMat observedDescriptors = new UMat();
surfCPU.DetectAndCompute(slicon, null, modelKeyPoints, modelDescriptors, false);
surfCPU.DetectAndCompute(test, null, observedKeyPoints, observedDescriptors, false);
var indices = new Matrix <int>(observedDescriptors.Rows, 2);
var dists = new Matrix <float>(observedDescriptors.Rows, 2);
var flannIndex = new Index(modelDescriptors, new KMeansIndexParams());
flannIndex.KnnSearch(observedDescriptors, indices, dists, 2);
for (int i = 0; i < indices.Rows; i++)
{
if (dists.Data[i, 0] < (0.6 * dists.Data[i, 1]))
{
int idx1 = indices[i, 0];
int idx2 = indices[i, 1];
Program.logIt(string.Format("{0}-{1}", indices[i, 0], indices[i, 1]));
MKeyPoint p1 = modelKeyPoints[idx1];
MKeyPoint p2 = observedKeyPoints[idx2];
Program.logIt(string.Format("{0}-{1}", p1.Point, p2.Point));
}
}
}
public static MKeyPoint[] toKeyPoints(this Vector2[] p)
{
MKeyPoint[] result = new MKeyPoint[p.Length];
for (int i = 0; i < p.Length; i++)
{
result[i].Point.X = p[i].x;
result[i].Point.Y = p[i].y;
result[i].Octave = 0;
result[i].Size = 2;
result[i].Angle = 0;
result[i].Response = 1;
}
return(result);
}
public void SetFeaturesVector(IEnumerable <uint> value)
{
KazePoints.Clear();
for (int i = 0; i < value.Count() / 7; i++)
{
MKeyPoint keyPoint = new MKeyPoint {
Angle = value.ElementAt(i + 0),
ClassId = (int)value.ElementAt(i + 1),
Octave = (int)value.ElementAt(i + 2),
Point = new System.Drawing.PointF {
X = value.ElementAt(i + 3),
Y = value.ElementAt(i + 4),
},
Response = value.ElementAt(i + 5),
Size = value.ElementAt(i + 6),
};
KazePoints.Add(keyPoint);
}
}
public void TestThreeViewsCorrespondences()
{
MKeyPoint[] kps1 = new MKeyPoint[]
{
new MKeyPoint()
{
Point = new PointF(0, 0)
},
new MKeyPoint()
{
Point = new PointF(5, 2)
},
new MKeyPoint()
{
Point = new PointF(5, 7)
},
new MKeyPoint()
{
Point = new PointF(2, 1)
},
new MKeyPoint()
{
Point = new PointF(-4, 2)
},
};
MKeyPoint[] kps2a = new MKeyPoint[]
{
new MKeyPoint()
{
Point = new PointF(0, 2)
},
new MKeyPoint()
{
Point = new PointF(9, 0)
},
new MKeyPoint()
{
Point = new PointF(9, -4)
},
new MKeyPoint()
{
Point = new PointF(6, 1)
},
new MKeyPoint()
{
Point = new PointF(-3, -3)
},
};
MKeyPoint[] kps2b = new MKeyPoint[]
{
new MKeyPoint()
{
Point = new PointF(0, 2)
},
new MKeyPoint()
{
Point = new PointF(9, 0)
},
new MKeyPoint()
{
Point = new PointF(9, -4)
},
new MKeyPoint()
{
Point = new PointF(6, 1)
},
new MKeyPoint()
{
Point = new PointF(-3, -3)
},
};
MKeyPoint[] kps3 = new MKeyPoint[]
{
new MKeyPoint()
{
Point = new PointF(0, 4)
},
new MKeyPoint()
{
Point = new PointF(11, 2)
},
new MKeyPoint()
{
Point = new PointF(11, -6)
},
new MKeyPoint()
{
Point = new PointF(8, 1)
},
new MKeyPoint()
{
Point = new PointF(-7, 0)
},
};
MDMatch[] matches12 = new MDMatch[4]
{
new MDMatch {
QueryIdx = 1, TrainIdx = 0
},
new MDMatch {
QueryIdx = 0, TrainIdx = 3
},
new MDMatch {
QueryIdx = 2, TrainIdx = 1
},
new MDMatch {
QueryIdx = 4, TrainIdx = 2
},
};
MDMatch[] matches23 = new MDMatch[4]
{
new MDMatch {
QueryIdx = 0, TrainIdx = 4
},
new MDMatch {
QueryIdx = 1, TrainIdx = 2
},
new MDMatch {
QueryIdx = 4, TrainIdx = 0
},
new MDMatch {
QueryIdx = 2, TrainIdx = 1
},
};
MatchingResult match12 = new MatchingResult()
{
LeftKps = kps1,
RightKps = kps2a,
Matches = new Emgu.CV.Util.VectorOfDMatch(matches12)
};
MatchingResult match23 = new MatchingResult()
{
LeftKps = kps2b,
RightKps = kps3,
Matches = new Emgu.CV.Util.VectorOfDMatch(matches23)
};
List <Correspondences.MatchPair> correspondences = Correspondences.FindCorrespondences12to23(match12, match23);
Assert.AreEqual(3, correspondences.Count);
// They are sorted same as matches23
var c = correspondences[0];
Assert.AreEqual(1, c.Match12.QueryIdx);
Assert.AreEqual(0, c.Match12.TrainIdx);
Assert.AreEqual(0, c.Match23.QueryIdx);
Assert.AreEqual(4, c.Match23.TrainIdx);
Assert.AreEqual(kps1[1].Point, c.Kp1.Point);
Assert.AreEqual(kps2a[0].Point, c.Kp2.Point);
Assert.AreEqual(kps3[4].Point, c.Kp3.Point);
c = correspondences[1];
Assert.AreEqual(2, c.Match12.QueryIdx);
Assert.AreEqual(1, c.Match12.TrainIdx);
Assert.AreEqual(1, c.Match23.QueryIdx);
Assert.AreEqual(2, c.Match23.TrainIdx);
c = correspondences[2];
Assert.AreEqual(4, c.Match12.QueryIdx);
Assert.AreEqual(2, c.Match12.TrainIdx);
Assert.AreEqual(2, c.Match23.QueryIdx);
Assert.AreEqual(1, c.Match23.TrainIdx);
}
public void TestMatchClosePoints()
{
Image <Arthmetic, double> desc1 = new Image <Arthmetic, double>(new double[, , ] {
{ { 2 }, { 2 }, },
{ { 1 }, { 0 }, },
{ { 2 }, { 0 }, },
{ { 0 }, { 1 }, },
{ { 1 }, { 1 }, },
{ { 0 }, { 2 }, },
});
Image <Arthmetic, double> desc2 = new Image <Arthmetic, double>(new double[, , ] {
{ { 1 }, { 0 }, },
{ { 1 }, { 1 }, },
{ { 0 }, { 1 }, },
{ { 2 }, { 0 }, },
{ { 2 }, { 2 }, },
{ { 0 }, { 2 }, },
});
MKeyPoint[] kps1 = new MKeyPoint[]
{
new MKeyPoint()
{
Point = new PointF(15, 0)
},
new MKeyPoint()
{
Point = new PointF(0, 0)
},
new MKeyPoint()
{
Point = new PointF(6, -5)
},
new MKeyPoint()
{
Point = new PointF(3, 0)
},
new MKeyPoint()
{
Point = new PointF(12, 0)
},
new MKeyPoint()
{
Point = new PointF(9, 0)
},
};
MKeyPoint[] kps2 = new MKeyPoint[]
{
new MKeyPoint()
{
Point = new PointF(0, 2)
},
new MKeyPoint()
{
Point = new PointF(9, 0)
},
new MKeyPoint()
{
Point = new PointF(3, 5)
},
new MKeyPoint()
{
Point = new PointF(6, 7)
},
new MKeyPoint()
{
Point = new PointF(15, 0)
},
new MKeyPoint()
{
Point = new PointF(9, 7)
},
};
var matches = MatchClosePoints.Match(kps1, kps2, desc1.Mat, desc2.Mat, Emgu.CV.Features2D.DistanceType.L2, 5.5, true);
}
List <float> ExtractSiftFeatureVector(TaggedImage image, int keyPointCount, SiftSortingMethod sortingMethod, bool doDrawImage)
{
// use the emgu functions to gather keypoints
VectorOfKeyPoint vectorOfKeypoints = new VectorOfKeyPoint();
Mat output = image.GetMat().Clone(); // only needed for drawing
sift.DetectAndCompute(image.GetMat(), null, vectorOfKeypoints, output, false);
// put it into useful data formats
List <MKeyPoint> keyPoints = new List <MKeyPoint>(vectorOfKeypoints.ToArray());
// sort
switch (sortingMethod)
{
case SiftSortingMethod.Response:
keyPoints.Sort((p1, p2) => p1.Response < p2.Response ? 1 : (p1.Response == p2.Response ? 0 : -1));
break;
case SiftSortingMethod.Size:
keyPoints.Sort((p1, p2) => p1.Size < p2.Size ? 1 : (p1.Size == p2.Size ? 0 : -1));
break;
case SiftSortingMethod.None:
default:
break;
}
// expand/trim
while (keyPoints.Count < keyPointCount)
{
keyPoints.Add(new MKeyPoint());
}
if (keyPoints.Count > keyPointCount)
{
keyPoints.RemoveRange(keyPointCount, keyPoints.Count - keyPointCount);
}
// visualize
if (doDrawImage)
{
vectorOfKeypoints = new VectorOfKeyPoint(keyPoints.ToArray());
Features2DToolbox.DrawKeypoints(image.GetMat(), vectorOfKeypoints, output, new Bgr(0, 0, 255), Features2DToolbox.KeypointDrawType.DrawRichKeypoints);
String win1 = "SIFT"; //The name of the window
CvInvoke.NamedWindow(win1); //Create the window using the specific name
CvInvoke.Imshow(win1, output); //Show the image
CvInvoke.WaitKey(0); //Wait for the key pressing event
CvInvoke.DestroyWindow(win1); //Destroy the window if key is pressed
}
// convert to list
List <float> result = new List <float>(5 * keyPointCount);
for (int i = 0; i < keyPoints.Count; ++i)
{
MKeyPoint current = keyPoints[i];
result.Add(current.Point.X / (float)output.Size.Width);
result.Add(current.Point.Y / (float)output.Size.Height);
result.Add(current.Size);
result.Add(current.Angle);
result.Add(current.Response);
}
return(result);
}
public ImagePointContext(Image image, MKeyPoint keyPoint)
{
Image = image;
KeyPoint = keyPoint;
}
public Vector2 NormalizePoint(MKeyPoint p)
{
return(new Vector2(p.Point.X / (float)_width, p.Point.Y / (float)_height));
}
internal static extern void VectorOfKeyPointGetItem(IntPtr vec, int index, ref MKeyPoint element);
internal static extern void VectorOfKeyPointGetItem(IntPtr keypoints, int index, ref MKeyPoint keypoint);
private List <MDMatch> clusterBasedOnPoseEstimation(List <MDMatch> matches, KeyPoints queryKeyPoints, PreProcessedImage trainingData)
{
// If no training data is provided then we cannot compute pose (LeafAnalysisV1), hence just return back the original set
if (trainingData == null ||
matches == null ||
!matches.Any())
{
return(matches);
}
Dictionary <MDMatch, List <MDMatch> > clusters = new Dictionary <MDMatch, List <MDMatch> >(matches.Count);
List <PoseEstimate> poseEstimates = new List <PoseEstimate>(matches.Count);
foreach (MDMatch match in matches)
{
MKeyPoint queryKeyPoint = queryKeyPoints.Points[match.QueryIdx];
MKeyPoint trainingKeyPoint = trainingData.KeyPoints.Points[match.TrainIdx];
PoseEstimate estimate = new PoseEstimate();
estimate.Match = match;
estimate.Dx = trainingKeyPoint.Point.X - queryKeyPoint.Point.X;
estimate.Dy = trainingKeyPoint.Point.Y - queryKeyPoint.Point.Y;
estimate.Ds = trainingKeyPoint.Octave / queryKeyPoint.Octave;
estimate.Do = trainingKeyPoint.Angle - queryKeyPoint.Angle;
poseEstimates.Add(estimate);
// Initialize clusters for each individual match
// Next we will add other matches which belong to this cluster
clusters.Add(match, new List <MDMatch>(new MDMatch[] { match }));
}
const double errorThreshold = 5;
// Compute cluster membership
foreach (PoseEstimate estimate in poseEstimates)
{
foreach (PoseEstimate otherEstimate in poseEstimates)
{
// Ignore self
if (estimate == otherEstimate)
{
continue;
}
double error = estimate.RMSE(otherEstimate);
//Console.WriteLine("Error: " + trainingData.Category + ": " + error);
if (error < errorThreshold)
{
clusters[estimate.Match].Add(otherEstimate.Match);
}
}
}
// Finally pick the largest cluster
List <MDMatch> result = null;
int sizeOfCluster = -1;;
foreach (KeyValuePair <MDMatch, List <MDMatch> > cluster in clusters)
{
if (cluster.Value.Count == sizeOfCluster)
{
// Tie breaker: choose the cluster with smaller overall distances
if (result.Sum(item => item.Distance) > cluster.Value.Sum(item => item.Distance))
{
result = cluster.Value;
}
}
else if (cluster.Value.Count > sizeOfCluster)
{
sizeOfCluster = cluster.Value.Count;
result = cluster.Value;
}
}
return(result);
}
/// <summary>
/// Identify good matches using RANSAC
/// </summary>/// symmetrical matches
/// keypoint1
/// keypoint2
/// the number of symmetrical matches
Matrix <double> ApplyRANSAC(Matrix <float> matches, VectorOfKeyPoint keyPoints1, VectorOfKeyPoint keyPoints2, int matchesNumber)
{
selPoints1 = new Matrix <float>(matchesNumber, 2);
selPoints2 = new Matrix <float>(matchesNumber, 2);
int selPointsIndex = 0;
for (int i = 0; i < matches.Rows; i++)
{
if (matches[i, 0] == 0 && matches[i, 1] == 0)
{
continue;
}
//Get the position of left keypoints
float x = keyPoints1[(int)matches[i, 0]].Point.X;
float y = keyPoints1[(int)matches[i, 0]].Point.Y;
selPoints1[selPointsIndex, 0] = x;
selPoints1[selPointsIndex, 1] = y;
//Get the position of right keypoints
x = keyPoints2[(int)matches[i, 1]].Point.X;
y = keyPoints2[(int)matches[i, 1]].Point.Y;
selPoints2[selPointsIndex, 0] = x;
selPoints2[selPointsIndex, 1] = y;
selPointsIndex++;
}
Matrix <double> fundamentalMatrix = new Matrix <double>(3, 3);
//IntPtr status = CvInvoke.cvCreateMat(1, matchesNumber, MAT_DEPTH.CV_8U);
//Matrix<double> status = new Matrix<double>(1, matchesNumber);
IntPtr statusp = CvInvoke.cvCreateMat(1, matchesNumber, MAT_DEPTH.CV_8U);
IntPtr points1 = CreatePointListPointer(selPoints1);
IntPtr points2 = CreatePointListPointer(selPoints2);
//IntPtr fundamentalMatrixp = CvInvoke.cvCreateMat(3, 3, MAT_DEPTH.CV_32F);
//Compute F matrix from RANSAC matches
CvInvoke.cvFindFundamentalMat(
points1, //selPoints1 points in first image
points2, //selPoints2 points in second image
fundamentalMatrix, //fundamental matrix
CV_FM.CV_FM_RANSAC, //RANSAC method
this._Distance, //Use 3.0 for default. The parameter is used for RANSAC method only.
this._Confidence, //Use 0.99 for default. The parameter is used for RANSAC or LMedS methods only.
statusp); //The array is computed only in RANSAC and LMedS methods.
Matrix <int> status = new Matrix <int>(1, matchesNumber, statusp);
//Matrix<double> fundamentalMatrix = new Matrix<double>(3, 3, fundamentalMatrixp);
if (this._RefineF)
{
matchesNumber = 0;
for (int i = 0; i < status.Cols; i++)
{
if (status[0, i] >= 1) // ==1
{
matchesNumber++;
}
}
selPoints1 = new Matrix <float>(matchesNumber, 2);
selPoints2 = new Matrix <float>(matchesNumber, 2);
modelKeyPoints = new VectorOfKeyPoint();
observedKeyPoints = new VectorOfKeyPoint();
int statusIndex = -1;
selPointsIndex = 0;
for (int i = 0; i < matches.Rows; i++)
{
if (matches[i, 0] == 0 && matches[i, 1] == 0)
{
continue;
}
statusIndex++;
if (status[0, statusIndex] >= 1) // == 1
{
//Get the position of left keypoints
float x = keyPoints1[(int)matches[i, 0]].Point.X;
float y = keyPoints1[(int)matches[i, 0]].Point.Y;
selPoints1[selPointsIndex, 0] = x;
selPoints1[selPointsIndex, 1] = y;
MKeyPoint[] kpt = new MKeyPoint[1];
kpt[0] = new MKeyPoint();
kpt[0].Point.X = x; kpt[0].Point.Y = y;
modelKeyPoints.Push(kpt);
//Get the position of right keypoints
x = keyPoints2[(int)matches[i, 1]].Point.X;
y = keyPoints2[(int)matches[i, 1]].Point.Y;
selPoints2[selPointsIndex, 0] = x;
selPoints2[selPointsIndex, 1] = y;
MKeyPoint[] kpt2 = new MKeyPoint[1];
kpt2[0] = new MKeyPoint();
kpt2[0].Point.X = x; kpt2[0].Point.Y = y;
observedKeyPoints.Push(kpt2);
selPointsIndex++;
}
}
status = new Matrix <int>(1, matchesNumber);
mask = new Matrix <byte>(matchesNumber, 1);
for (int i = 0; i < mask.Rows; i++)
{
mask[i, 0] = 0; // don't draw lines, we will do it our selves
} // set this to one if you wanted to use Features2DToolbox.DrawMatches
// to draw correspondences
indices = new Matrix <int>(matchesNumber, 2); // not being used as we draw correspondences
for (int i = 0; i < indices.Rows; i++) // ourselves
{
indices[i, 0] = i; // has a problem in drawing lines, so we will drawe ourselves
indices[i, 1] = i; // this is not being used in our code
}
//Compute F matrix from RANSAC matches // we can do additional RANSAC filtering
//CvInvoke.cvFindFundamentalMat( // but first RANSAC gives good results so not used
// selPoints1, //points in first image
// selPoints2, //points in second image
// fundamentalMatrix, //fundamental matrix
// CV_FM.CV_FM_RANSAC, //RANSAC method
// this._Distance, //Use 3.0 for default. The parameter is used for RANSAC method only.
// this._Confidence, //Use 0.99 for default. The parameter is used for RANSAC or LMedS methods only.
// status);//The array is computed only in RANSAC and LMedS methods.
// we will need to copy points from selPoints1 and 2 based on status if above was uncommented
}
return(fundamentalMatrix);
}
internal static extern void VectorOfKeyPointGetItem(IntPtr keypoints, int index, ref MKeyPoint keypoint);
public static String ToString(MKeyPoint keyPoint, String indent = "")
{
return(indent + JsonConvert.SerializeObject(keyPoint));
}
public KeyPoint(MKeyPoint pixel)
{
_pixel = pixel;
}
