Robust image hashing against geometric attacks using image normalization and SVD decomposition

摘要

This paper uses SVD decomposition and image normalization algorithm to generate a hash value of a digital image. SVD decomposition is an important linear algebra tool, which is often used in image compression, digital watermarking and other signal processing fields, on the other hand image normalization is an algorithm that has been shown to be robust to different kinds of affine transforms. Combining these two techniques we can generate a hash value which is robust against geometric attacks. The proposed algorithm is shown to be robust against some common attacks such as JPEG compression, rotation, content corruption and cropping.