Rect haarPatternDetection(CascadeClassifier classifier, Mat image, int imageWidthforDetection, Rect parentRect=Rect()) {
double imageSizeScale = 1.0*image.size().width*1.0/image.size().height;
Mat color_img = image.clone();
Mat face_img, gray_img;
imresize(color_img, imageWidthforDetection,face_img);
cvtColor(face_img, gray_img, CV_BGR2GRAY);
equalizeHist(gray_img, gray_img);
vector<Rect> faceRectsVector;
classifier.detectMultiScale(gray_img, faceRectsVector,1.2,3,
CV_HAAR_FIND_BIGGEST_OBJECT | CV_HAAR_SCALE_IMAGE );
Rect faceRect(0,0,0,0);
for (int i = 0; i < faceRectsVector.size(); i++) {
drawBox(face_img, faceRectsVector[i]);
if (faceRect.area() < faceRectsVector[i].area())
faceRect = faceRectsVector[i];
}
drawBox(face_img, faceRect);
// imshow("rect",face_img);
double x = 1.0*faceRect.x/(double)imageWidthforDetection;
double y = 1.0*faceRect.y/(double)cvRound(imageWidthforDetection/imageSizeScale);
double w = 1.0*faceRect.width/(double)imageWidthforDetection;
double h = 1.0*faceRect.height/(double)cvRound(imageWidthforDetection/imageSizeScale);
int bx = cvRound(x*image.size().width) ;;//+ parentRect.tl().x;
int by = cvRound(y*image.size().height) ;//+ parentRect.tl().y;
int bw = cvRound(w*image.size().width);
int bh = cvRound(h*image.size().height);
Rect biggerRect(bx,by,bw,bh);
biggerRect += parentRect.tl();
return biggerRect;
}
void Slic::build_pyramid(const vector<unsigned char>& img, int levels) {
#ifdef SLIC_DEBUG
unsigned long t1, t2;
t1 = now_us();
printf("Slic::build_pyramid ");
#endif
// build the pyramid for the image data
pyramid = vector<vector<unsigned char> >(levels);
int n = rows[0] * cols[0] * chan;
pyramid[0] = vector<unsigned char>(n);
copy(img.begin(), img.begin() + n, pyramid[0].begin());
for (int i = 1; i < levels; ++i) {
imresize(pyramid[i], pyramid[i - 1], rows[i - 1], cols[i - 1], rows[i], cols[i]);
}
// build the pyramid for assignments:
assignments_pyramid = vector<vector<int> >(levels);
for (int i = 0; i < levels; ++i) {
assignments_pyramid[i] = vector<int>(rows[i] * cols[i], 0);
}
#ifdef SLIC_DEBUG
t2 = now_us();
printf("elapsed: %f\n", (t2 - t1) / 1000.0);
#endif
}
void MyProcessingClass::ResizeImage()
{
/*
* *
* function [newIm,zoom]=ResizeImage(Im,oldSizeOfROI,newSizeOfROI)
zoom = newSizeOfROI/oldSizeOfROI;
[height,width,deep] = size(Im);
newIm = imresize(Im,[height*zoom width*zoom]);
end
*/
float zoom = m_new_size_of_roi / m_old_size_of_roi;
size(m_input_image, m_image_size, m_image_depth);
m_input_image = imresize(m_input_image,Sizei(m_input_image.size().height * zoom, m_input_image.size().width * zoom));
}
void findEyeCenterByColorSegmentation(const Mat& image, Point2f & eyeCoord, float coordinateWeight, int kmeansIterations, int kmeansRepeats, int blurSize) {
Mat img, gray_img;
Mat colorpoints, kmeansPoints;
img = equalizeImage(image);
medianBlur(img, img, blurSize);
cvtColor(image, gray_img, CV_BGR2GRAY);
gray_img = imcomplement(gray_img);
vector<Mat> layers(3);
split(img, layers);
for (int i = 0 ; i < layers.size(); i++) {
layers[i] = layers[i].reshape(1,1).t();
}
hconcat(layers, colorpoints);
// add coordinates
colorpoints.convertTo(colorpoints, CV_32FC1);
Mat coordinates = matrixPointCoordinates(img.rows,img.cols,false) *coordinateWeight;
hconcat(colorpoints, coordinates, kmeansPoints);
Mat locIndex(img.size().area(),kmeansIterations,CV_32FC1,Scalar::all(-1));
linspace(0, img.size().area(), 1).copyTo(locIndex.col(0));
Mat index_img(img.rows,img.cols,CV_32FC1,Scalar::all(0));
Mat bestLabels, centers, clustered , colorsum , minColorPtIndex;
for(int it = 1 ; it < kmeansIterations ; it++) {
if (kmeansPoints.rows < 2) {
break;
}
kmeans(kmeansPoints,2,bestLabels,TermCriteria(CV_TERMCRIT_EPS+CV_TERMCRIT_ITER, kmeansRepeats, 0.001),kmeansRepeats,KMEANS_PP_CENTERS,centers);
reduce(centers.colRange(0, 3), colorsum, 1, CV_REDUCE_SUM);
if (colorsum.at<float>(0) < colorsum.at<float>(1)) {
findNonZero(bestLabels==0, minColorPtIndex);
}
else {
findNonZero(bestLabels==1, minColorPtIndex);
}
minColorPtIndex = minColorPtIndex.reshape(1).col(1);
for (int i = 0; i <minColorPtIndex.rows ; i ++) {
locIndex.at<float>(i,it) = locIndex.at<float>(minColorPtIndex.at<int>(i),it-1);
}
Mat temp;
for (int i = 0; i <minColorPtIndex.rows ; i ++) {
temp.push_back(kmeansPoints.row(minColorPtIndex.at<int>(i)));
}
temp.copyTo(kmeansPoints);
temp.release();
for (int i = 0 ; i < minColorPtIndex.rows ; i ++) {
int r, c;
ind2sub(locIndex.at<float>(i,it), index_img.cols, index_img.rows, r, c);
index_img.at<float>(r,c) +=1;
}
}
// imagesc("layered",mat2gray(index_img));
Mat layerweighted_img = index_img.mul(index_img);
layerweighted_img = mat2gray(layerweighted_img);
gray_img.convertTo(gray_img, CV_32FC1,1/255.0);
Mat composed = gray_img.mul(layerweighted_img);
float zoomRatio = 5.0f;
Mat zoomed;
imresize(composed, zoomRatio, zoomed);
Mat score = calculateImageSymmetryScore(zoomed);
// imagesc("score", score);
Mat scoresum;
reduce(score.rowRange(0, zoomed.cols/6), scoresum, 0, CV_REDUCE_SUM,CV_32FC1);
// plotVectors("scoresum", scoresum.t());
double minVal , maxVal;
Point minLoc, maxLoc;
minMaxLoc(scoresum,&minVal,&maxVal,&minLoc,&maxLoc);
float initialHC = (float)maxLoc.x/zoomRatio;
line(zoomed, Point(maxLoc.x,0), Point(maxLoc.x,zoomed.rows-1), Scalar::all(255));
// imshow("zoomed", zoomed);
int bestx = 0,bestlayer = 0;
Mat bestIndex_img = index_img >=1;
minMaxLoc(index_img,&minVal,&maxVal,&minLoc,&maxLoc);
for (int i = 1 ; i<=maxVal; i++) {
Mat indexlayer_img = index_img >=i;
medianBlur(indexlayer_img, indexlayer_img, 5);
erode(indexlayer_img, indexlayer_img, blurSize);
erode(indexlayer_img, indexlayer_img, blurSize);
indexlayer_img = removeSmallBlobs(indexlayer_img);
indexlayer_img = fillHoleInBinary(indexlayer_img);
indexlayer_img = fillConvexHulls(indexlayer_img);
Mat score = calculateImageSymmetryScore(indexlayer_img);
Mat scoresum;
reduce(score.rowRange(0, indexlayer_img.cols/6), scoresum, 0, CV_REDUCE_SUM,CV_32FC1);
minMaxLoc(scoresum,&minVal,&maxVal,&minLoc,&maxLoc);
if (abs(maxLoc.x - initialHC) < abs(bestx - initialHC)) {
if (sum(indexlayer_img)[0]/255 < indexlayer_img.size().area()/5*2 &&
sum(indexlayer_img)[0]/255 > indexlayer_img.size().area()/6) {
bestx = maxLoc.x;
bestlayer = i;
bestIndex_img = indexlayer_img.clone();
}
}
}
Point2f massCenter = findMassCenter_BinaryBiggestBlob(bestIndex_img);
eyeCoord = Point2f(initialHC,massCenter.y);
}
Masek::IMAGE* Masek::canny(IMAGE *im, double sigma, double scaling,
double vert, double horz, filter *gradient, filter *orND)
{
filter gaussian, *newim;
int i, j;
int hsize[2];
int rows, cols;
double xscaling, yscaling;
double *h, *v, *d1, *d2, X, Y, begin, end;
xscaling = vert;
yscaling = horz;
hsize[0] = (int)(6*sigma+1);
hsize[1] = (int)(6*sigma+1); // % The filter size.
gaussian.hsize[0] = hsize[0];
gaussian.hsize[1] = hsize[1];
gaussian.data = (double*) malloc(sizeof(double)*hsize[0]*hsize[1]);
/*gaussian = fspecial('gaussian',hsize,sigma);
im = filter2(gaussian,im); % Smoothed image.*/
CREATEGAUSS (hsize, sigma, &gaussian);
newim = filter2(gaussian,im);
newim = imresize(newim, scaling);
rows = newim->hsize[0];
cols = newim->hsize[1];
h = (double*)malloc(sizeof(double)*rows*cols);
for (i = 0; i<rows; i++)
{
for (j = 0; j<cols; j++)
{
if (j == 0)
*(h+i*cols+j) = (newim->data[i*cols+1]);
else if (j == cols-1)
*(h+i*cols+j) = (-newim->data[i*cols+j-1]);
else
*(h+i*cols+j) = (newim->data[i*cols+j+1]-newim->data[i*cols+j-1]);
}
}
v = (double*)malloc(sizeof(double)*rows*cols);
for (i = 0; i<rows; i++)
{
for (j = 0; j<cols; j++)
{
if (i == 0)
*(v+i*cols+j) = (newim->data[(i+1)*cols+j]);
else if (i == rows-1)
*(v+i*cols+j) = -newim->data[(i-1)*cols+j];
else
*(v+i*cols+j) = (newim->data[(i+1)*cols+j]-newim->data[(i-1)*cols+j]);
}
}
d1 = (double*)malloc(sizeof(double)*rows*cols);
d2 = (double*)malloc(sizeof(double)*rows*cols);
for (i = 0; i<rows; i++)
{
for (j = 0; j<cols; j++)
{
if (i == rows-1 || j == cols-1)
begin = 0;
else
begin = newim->data[(i+1)*cols+j+1];
if (i == 0 || j == 0)
end = 0;
else
end = newim->data[(i-1)*cols+j-1];
*(d1+i*cols+j) = begin-end;
}
}
for (i = 0; i<rows; i++)
{
for (j = 0; j<cols; j++)
{
if (i == 0 || j == cols-1)
begin = 0;
else
begin = newim->data[(i-1)*cols+j+1];
if (i == rows-1 || j == 0)
end = 0;
else
end = newim->data[(i+1)*cols+j-1];
*(d2+i*cols+j) = begin-end;
}
}
orND->data = (double*)malloc(sizeof(double)*rows*cols);
orND->hsize[0] = rows;
orND->hsize[1] = cols;
gradient->data = (double*)malloc(sizeof(double)*rows*cols);
gradient->hsize[0] = rows;
gradient->hsize[1] = cols;
for (i = 0; i<rows*cols;i++)
{
X = (h[i]+(d1[i]+d2[i])/2.0)*xscaling;
Y = (v[i]+(d1[i]-d2[i])/2.0)*yscaling;
gradient->data[i] = sqrt(X*X+Y*Y);
orND->data[i] = atan2(-Y, X);
/*if (i == 14)
{
printf("h is %f, d1 is %f, d2 is %f, v is %f xscaling is %f, yscaling is %f\n", h[i], d1[i], d2[i], v[i], xscaling, yscaling);
printf("X is %f, Y is %f\n", X, Y);
printf("orND is %f\n", orND->data[i]);
}
*/
if (orND->data[i]<0)
orND->data[i]+=PI;
// if (i == 14)
// printf("orND is %f\n", orND->data[i]);
orND->data[i] = (orND->data[i]/PI)*180;
// if (i == 14)
// printf("orND is %f\n", orND->data[i]);
}
free (gaussian.data);
free(d1);
free(d2);
free(h);
free(v);
free(newim->data);
free(newim);
return im;
}