Quantifying stitching effects

Question

I have many images like this with high stitching, medium stitching, low stitching, and no stitching effects (the adjectives high, medium, low and no are purely subjective). How to quantify the amount of stitching that happens in each image since I want to eliminate images with high stitching? Thanks.

P.S We tried Fourier transform to isolate high frequency components but it doesn't seem to work.

Answer 1

You could try using an edge detector with different thresholds, the more horizontal and vertical edges you will find, the more stitches in your overall image.

For guidance this is where I would start with implementing, using opencv.

To be able to answer this question in more depth, as mentioned in a comment, we would need sample data - even it is some artificially generated one.