Open carcinocron opened 3 years ago
The duplicatefinder app considers a file as a duplicate if they are binary equal. This equality is proven by comparing their SHA-256 hashes.
For a semantical comparison of images, there already exists a feature request #10
Thanks a lot for this clarification! I think it would be very helpful to include this in the Readme, pointing out, that duplicates are not found by filename, path, creation-timestamp or anything else "superficial". Maybe even a link to Wikipedia -> SHA256 would be helpful, pointing out the de facto impossibility of wrong matches.
Are duplicates done by md5/sha hash? perceptual hash? Does it only detect literal identical files or does it help find files that are the same photo but one is a smaller resolution or cropped?