Hi,
I'm looking for an algorithm to normalize images to be able to compare them with QualitySSIM.compute() method.
I've created the below code which works fine but just in case one picture is small than the second,
but in case one is shorter and one is narrower the program failed.
Mat result = new Mat();
Mat resultMask = new Mat();
opencv_imgproc.matchTemplate(base, screenShot, result, opencv_imgproc.TM_CCORR);
double[] min = new double[2], max = new double[2];
Point minLoc = new Point(), maxLoc = new Point();
opencv_core.minMaxLoc(result, min, max, minLoc, maxLoc,resultMask);
Rect r = new Rect(maxLoc, new Point(maxLoc.x() + 619, maxLoc.y() + 713));
scrnShtForCompare = new Mat(bigMat, r);
}
Scalar ssim = QualitySSIM.compute(base,screenshot, diffImage);
double matchingPercentages=( (width+height) /(base.cols()+base.rows())) * ssim.get(0);
Any help??