I'm trying to find the best algorithm for my task. I have images that look like this:
I want to find the most similar from a set of such images. I've tried several algorithms to calculate the similarity value (sieves – opencv2, structural similarities – skimage.measure.compare_ssim, pixel similarity, wasserstein_distance …). None of them seems to work as I would like. For example, sift returned an almost maximal similarity for the image above and this one:
Which is far from perfect. Are there any algorithms that could better perform this task?