#Proposal This project would implement an analysis program that would identify between two images of they have any likelyhood of being similar or the same. It is not to find exactly identical images with same dimension, same color space, metadata, or file formating. The program should be able to find similarity with different dimension, color space, or any augmentation that still retains an image pictoral properties. A sample list of augmentation is crop, invert, rotation, and blur. The program will identify the differences between two image, give back a result of the differences in percentage. #Use case This would be use to find duplicate or very similar image in order to remove or reorganize a person's image collection.
My use case is to find duplicate comics within my database. Since there are comics that are single chapter and another comics that could contain that chapter, this can be use to identify the that aspect of duplication in a collection. The implement of pdf extraction and and matching can if I would have multiple documents of something in order to clean for space. #Outcome The result will be the directory in a text file