Universal distance measure for images

摘要: We introduce an algorithm for measuring the distance between two images based on computing complexity of strings characters that encode images. Given a pair images, our transforms each one into text-based sequence (strings) characters. For string, it computes LZ-complexity and then uses string-distance measure [1] to obtain value The main advantages are is universal, is, neither needs nor assumes any spatial or spectral information about can different sizes, works black white, grayscale color be implemented efficiently embedded computer system. present successful experimental results clustering sizes categories their similarities as measured by algorithm.