public CvKNearest TrainData(IList <ImageInfo> trainingImages)
{
var samples = new Mat();
foreach (var trainingImage in trainingImages)
{
samples.PushBack(trainingImage.Image);
}
var labels = trainingImages.Select(x => x.ImageGroupId).ToArray();
var responses = new Mat(labels.Length, 1, MatType.CV_32SC1, labels);
var tmp = responses.Reshape(1, 1); //make continuous
var responseFloat = new Mat();
tmp.ConvertTo(responseFloat, MatType.CV_32FC1); // Convert to float
var kNearest = new CvKNearest();
kNearest.Train(samples, responseFloat); // Train with sample and responses
return(kNearest);
}
//invoke knn class
public void train()
{
knn = CvKNearest(trainData, trainClasses, false, K);
}
//invoke knn class
public void train()
{
knn = new CvKNearest(trainData, trainClasses, false, K);
}
public void DoOCR(CvKNearest kNearest, string path)
{
var src = Cv2.ImRead(path);
Cv2.ImShow("Source", src);
var gray = new Mat();
Cv2.CvtColor(src, gray, ColorConversion.BgrToGray);
var threshImage = new Mat();
Cv2.Threshold(gray, threshImage, Thresh, ThresholdMaxVal, ThresholdType.BinaryInv); // Threshold to find contour
Point[][] contours;
HiearchyIndex[] hierarchyIndexes;
Cv2.FindContours(
threshImage,
out contours,
out hierarchyIndexes,
mode: ContourRetrieval.CComp,
method: ContourChain.ApproxSimple);
if (contours.Length == 0)
{
throw new NotSupportedException("Couldn't find any object in the image.");
}
//Create input sample by contour finding and cropping
var dst = new Mat(src.Rows, src.Cols, MatType.CV_8UC3, Scalar.All(0));
var contourIndex = 0;
while ((contourIndex >= 0))
{
var contour = contours[contourIndex];
var boundingRect = Cv2.BoundingRect(contour); //Find bounding rect for each contour
Cv2.Rectangle(src,
new Point(boundingRect.X, boundingRect.Y),
new Point(boundingRect.X + boundingRect.Width, boundingRect.Y + boundingRect.Height),
new Scalar(0, 0, 255),
2);
var roi = new Mat(threshImage, boundingRect); //Crop the image
var resizedImage = new Mat();
var resizedImageFloat = new Mat();
Cv2.Resize(roi, resizedImage, new Size(10, 10)); //resize to 10X10
resizedImage.ConvertTo(resizedImageFloat, MatType.CV_32FC1); //convert to float
var result = resizedImageFloat.Reshape(1, 1);
var results = new Mat();
var neighborResponses = new Mat();
var dists = new Mat();
var detectedClass = (int)kNearest.FindNearest(result, 1, results, neighborResponses, dists);
//Console.WriteLine("DetectedClass: {0}", detectedClass);
//Cv2.ImShow("roi", roi);
//Cv.WaitKey(0);
//Cv2.ImWrite(string.Format("det_{0}_{1}.png",detectedClass, contourIndex), roi);
Cv2.PutText(
dst,
detectedClass.ToString(CultureInfo.InvariantCulture),
new Point(boundingRect.X, boundingRect.Y + boundingRect.Height),
0,
1,
new Scalar(0, 255, 0),
2);
contourIndex = hierarchyIndexes[contourIndex].Next;
}
Cv2.ImShow("Segmented Source", src);
Cv2.ImShow("Detected", dst);
Cv2.ImWrite("dest.jpg", dst);
Cv2.WaitKey();
}
