Adaptive local feature based multi-scale image hashing for robust tampering detection

This paper proposes a novel multi-scale image hashing method by using the location-context information of the features generated by adaptive local feature extraction techniques. The adaptive local feature extraction method is proposed for more robust feature descriptors. The global hash is calculated to determine whether the received image has been maliciously tampered. The multi-scale hash is calculated to locate the tampered regions. Experimental results show that the proposed tampering detection scheme is very robust against the content-preserving attacks, including both common signal processing and geometric distortions.