Ok, now I have a repository called Codecember where I have all the images that I’ve downloaded. Now It’s time to verify the similarities of the images. Googling a quick and fast way to do so. I want to avoid OpenCV since I’m probably running this script on a super light resource, I’ve found that the best way would be through hashing.

https://medium.com/@somilshah112/how-to-find-duplicate-or-similar-images-quickly-with-python-2d636af9452f

It seems to work very well on photos of people. Not working so well with photos of landscapes. Moreover, on ~ 5000 images, it found only 13 duplicated images.

Maybe I need a double check my work.

Follow my work on github! github.com/girolamodaschio/codecember