I have a lot of—hundreds of thousands—JPEG images, and most of them are very similar to each other. While they aren’t of high quality and take about 100+ kB each, this adds up to a lot of storage. Are there any options to compress them which take the similarity into account?
I thought about something like using a video compressor which should perform well, but the images—not all of them, of course, but blocks of several thousands sequential ones—are often needed as separate files for analysis, with correct file names. Also new images are added at a rate of about 100/hour. So, just converting them all to a video once doesn’t suffice. I suppose using Fuse can help somehow? There doesn’t seem to be a ready-to-use solution, so any suggestions are welcome.
Best Answer
To find similar images you can use software like SimilarImages: https://tn123.org/simimages/
or Image Comparer: http://image-comparer.software.informer.com/
or many others: http://similar-images.software.informer.com/
Generally speaking, JPEGS are already highly compressed images so compressing them with a standard WINRAR\WINZIP\7-ZIP program would give little to no further compression. I'd recommend what was once used in Windows XP, called imageresizer, from Microsoft Power Tools, is now available for Vista and 7, here: https://imageresizer.codeplex.com/
It does batch resizing and the image looks great even after the size is reduced to almost 10 times the original size.