Secure Image Verification in Jointed Frequency Domains

The paper illustrates an image hashing scheme and an image verification system based on jointed hashing and watermarking algorithms. We propose an image hashing scheme based on a set of robust features extracted from jointed DCT-DWT frequency domains and enciphered using an efficient chaotic system. The proposed hashing algorithm and a rotation-scale-translation (RST) invariant watermarking scheme are used to design a robust verification system for digital images. A robustness investigation is conducted using compression, noise addition, filtering and geometric transforms. The Checkmark benchmark is used to achieve most of the attacks. Our experiments reveal the performances of the hashing scheme and also the success of the proposed verification system against common image processing attacks.