Comment by kaivi

This is all good, but how does one store, index and efficiently search for near matches? I have a side project with terabytes of photo URLs, and hashing&indexing them has been an itch I could not scratch.