4 ms·

Out of curiosity, what methods did you use for image dedupes? Simple file hash, or something more complex to find e.g. watermark vs no, different crops/scales,

• drbawb 11y ago
  At the moment it is just hash based, which actually remedied most of my immediate problem. (I have 8 drives in my workstation. Many of them are donors from computers long gone. [Portions of] my image library have just been copied around many times without much rhyme or reason. Now it's centralized and properly backed up.) I did think about comparing images by similarity but ultimately tabled it for a later evening because: - I quickly realized I had a lot of reading to do - Resolving "similar images" conflicts is a fuzzier, more time intensive process since: (a) maybe the comparison is just plain wrong?, (b) maybe it's not strictly a worse crop, but rather two common aspect ratios which you'd like to keep, (c) maybe it's not actually "watermark" text but rather it's a meme or image macro, etc. - Personally: my real problem with images of varying sizes is that I download thumbnails by mistake. That's just PEBKAC and no amount of software can fix that. (Now that I think about it: it might be nice to have a script which iterates over low resolution images and tries to find a large resolution version w/ Google Image Search.)