private Double SkewAngle(IplImage img)
{
Double result = 0.0;
IplImage tempImage = null;
try
{
tempImage = new IplImage(img.Size, img.Depth, 1);
Cv.CvtColor(img, tempImage, ColorConversion.RgbaToGray);
//Cv.Not(tempImage, tempImage);
Mat imgMat = new Mat(tempImage);
CvCpp.Canny(imgMat, imgMat, 50, 200, ApertureSize.Size3, false);
CvLineSegmentPoint[] segProb = CvCpp.HoughLinesP(imgMat, 1, Math.PI / 180.0, 50, ((Double)img.Size.Width) / 2.0, 10.0);
foreach (CvLineSegmentPoint point in segProb)
{
result += Math.Atan2((Double)(point.P2.Y) - (Double)(point.P1.Y), (Double)(point.P2.X) - (Double)(point.P1.X));
//img.Line(point.P1,point.P2,new CvScalar(255,0,0));
}
if (segProb.Length > 0)
{
result /= (Double)(segProb.Length);
result = result * 180 / Math.PI;
}
}
finally
{
if (null != tempImage)
{
Cv.ReleaseImage(tempImage);
}
}
//img.SaveImage(@"C:\temp\newImages\res.jpg");
return(result);
}
/// <summary>
/// Try to recognize face on the sample image by fisherfaces algorithm
/// </summary>
/// <param name="err">Eror happened due the execution of the subroutine</param>
/// <param name="trainData">Array of images train a system</param>
/// <param name="sample">Face to recognize</param>
/// <param name="threshold">Threshold of the algorithm. Defines maximum length of the eigenvector length for recognition</param>
/// <param name="cropImage">Defines whether it is need to crop sample image (to find a face)</param>
/// <returns></returns>
public static bool ProcessImageSample(out string err, out int predictedLabel, Dictionary <int, List <byte[]> > trainData, byte[] sample, int threshold = 3500, bool cropImage = false)
{
IplImage sampleImg = null; //image to recognize
Mat eigenvalues = null; //eigenvalues
Mat W = null; //eigenvectors
Mat mean = null; //mean images
CvSize actualImagesSize; //size of train data and sample images
List <Mat> trainImages = new List <Mat>();
err = "";
predictedLabel = -1;//label for predict (user id)
List <int> labels = new List <int>();
try
{
if (sample == null)
{
throw new ArgumentException("sample == null");
}
sampleImg = support.ByteArrayToIplImage(sample, LoadMode.GrayScale);
if (sampleImg == null)
{
throw new Exception("Could not convert sample byte array to IplImage");
}
//Convert train data dictionary in two Lists (with labels and images)
foreach (KeyValuePair <int, List <byte[]> > ms in trainData)
{
foreach (byte[] m in ms.Value)
{
trainImages.Add(support.ByteArrayToIplImage(m, LoadMode.GrayScale));
labels.Add(ms.Key);
}
}
// Quit if there are not enough images for this demo.
if (trainImages.Count <= 1)
{
throw new Exception("Needs at least 2 images to work. Please add more images to your data set!");
}
else
{
actualImagesSize.Height = trainImages[0].Rows;
actualImagesSize.Width = trainImages[0].Cols;
}
// Get the height from the first image. We'll need this
// later in code to reshape the images to their original
// size:
int height = trainImages[0].Rows;
//Crop the face if need
if (cropImage)
{
IplImage[] cropImgs = HaarCascade.GetFaces(sampleImg);
if (cropImgs.Length > 0)
{
sampleImg = support.ResizeImage(cropImgs[0], actualImagesSize).ToIplImage();
}
foreach (IplImage im in cropImgs)
{
Cv.ReleaseImage(im);
}
}
if (sampleImg.ElemChannels > 1)
{
throw new Exception("sampleImg should be grayscale!");
}
// The following lines create an Fisherfaces model for
// face recognition and train it with the images and
// labels read from the given CSV file.
// If you just want to keep 10 Fisherfaces, then call
// the factory method like this:
//
// cv::createFisherFaceRecognizer(10);
//
// However it is not useful to discard Fisherfaces! Please
// always try to use _all_ available Fisherfaces for
// classification.
//
// If you want to create a FaceRecognizer with a
// confidence threshold (e.g. 123.0) and use _all_
// Fisherfaces, then call it with:
//
// cv::createFisherFaceRecognizer(0, 123.0);
//
predictedLabel = CvCpp.ProcessFisherFaceRecognizer(trainImages, labels, sampleImg, out eigenvalues, out W, out mean, 0, threshold);
if (predictedLabel < 0)
{
return(false);
}
Cv.ReleaseImage(sampleImg);
foreach (IplImage im in trainImages)
{
Cv.ReleaseImage(im);
}
return(true);
}
catch (System.Exception ex)
{
err = ex.Message;
return(false);
}
}
/// <summary>
/// Try to recognize face on the sample image by eigenfaces algorithm
/// </summary>
/// <param name="err">Eror happened due the execution of the subroutine</param>
/// <param name="trainData">Array of images train a system</param>
/// <param name="sample">Face to recognize</param>
/// <param name="threshold">Threshold of the algorithm. Defines maximum length of the eigenvector length for recognition</param>
/// <param name="cropImage">Defines whether it is need to crop sample image (to find a face)</param>
/// <returns></returns>
public static bool ProcessImageSample(out string err, out int predictedLabel, Dictionary <int, List <byte[]> > trainData, byte[] sample, int threshold = 3500, bool cropImage = false)
{
IplImage sampleImg = null; //image to recognize
Mat eigenvalues = null; //eigenvalues
Mat W = null; //eigenvectors
Mat mean = null; //mean images
CvSize actualImagesSize; //size of train data and sample images
List <Mat> trainImages = new List <Mat>();
err = "";
predictedLabel = -1;
List <int> labels = new List <int>();
try
{
if (sample == null)
{
throw new ArgumentException("sample == null");
}
sampleImg = support.ByteArrayToIplImage(sample, LoadMode.GrayScale);
if (sampleImg == null)
{
throw new Exception("Could not convert sample byte array to IplImage");
}
foreach (KeyValuePair <int, List <byte[]> > ms in trainData)
{
foreach (byte[] m in ms.Value)
{
trainImages.Add(support.ByteArrayToIplImage(m, LoadMode.GrayScale));
labels.Add(ms.Key);
}
}
// Quit if there are not enough images for this demo.
if (trainImages.Count <= 1)
{
throw new Exception("Needs at least 2 images to work. Please add more images to your data set!");
}
else
{
actualImagesSize.Height = trainImages[0].Rows;
actualImagesSize.Width = trainImages[0].Cols;
}
// Get the height from the first image. We'll need this
// later in code to reshape the images to their original
// size:
int height = trainImages[0].Rows;
// The following lines simply get the last images from
// your dataset and remove it from the vector. This is
// done, so that the training data (which we learn the
// cv::FaceRecognizer on) and the test data we test
// the model with, do not overlap.
if (cropImage)
{
IplImage[] cropImgs = HaarCascade.GetFaces(sampleImg);
if (cropImgs.Length > 0)
{
sampleImg = support.ResizeImage(cropImgs[0], actualImagesSize).ToIplImage();
}
foreach (IplImage im in cropImgs)
{
Cv.ReleaseImage(im);
}
}
if (sampleImg.ElemChannels > 1)
{
throw new Exception("sampleImg should be grayscale!");
}
// The following line create an Eigenfaces model for
// face recognition and train it with the images and
// This here is a full PCA, if you just want to keep
// 10 principal components (read Eigenfaces), then call
// the factory method like this:
//
// cv::createEigenFaceRecognizer(10);
//
// If you want to create a FaceRecognizer with a
// confidence threshold (e.g. 123.0), call it with:
//
// cv::createEigenFaceRecognizer(10, 123.0);
//
// If you want to use _all_ Eigenfaces and have a threshold,
// then call the method like this:
//
// cv::createEigenFaceRecognizer(0, 123.0);
//
predictedLabel = CvCpp.ProcessEigenFaceRecognizer(trainImages, labels, sampleImg, out eigenvalues, out W, out mean, 0, threshold);
if (predictedLabel < 0)
{
return(false);
}
Cv.ReleaseImage(sampleImg);
foreach (IplImage im in trainImages)
{
Cv.ReleaseImage(im);
}
return(true);
}
catch (System.Exception ex)
{
err = ex.Message;
return(false);
}
}
