Robust image hashing using image normalization and SVD decomposition

This paper introduces the use of image normalization and SVD decomposition hash functions to generate a hash value for a digital image. Image normalization is a technique that has been shown to be robust against different kinds of geometric attacks such as rotation, scaling and flipping. Using image normalization as a preprocessing step we can increase the robustness of the hash by 50% against rotation and scaling compared to without this algorithm.