DCT sign-based robust image hashing

Abstract

This paper proposes a robust image hashing technique based on the sign of coefficients in the Discrete Cosine Transforms (DCT) domain. The idea consists of exploiting the energy compaction property of the DCT and its ability to carry information of edges and texture in DCT sign values. It has been shown that the DCT sign-based hashing technique offers an excellent robustness against signal processing operations and geometric attacks and outperforms the conventional DCT-based image hashing technique.