Perceptual Hash Functions Based on Contourlet Transform and Singular Value Decomposition

Perceptual image hashing maps an image to a fixed length binary string based on the image\´s appearance to the human eye, and has applications in content authentication, watermarking and anti-piracy search. In this paper, we present a hash scheme that is robust and secure to non-malicious manipulation and sensitive to the malicious tampering. The algorithm is based on contourlet transform and singular value decomposition. Due to Contourlet is a "true" two dimensional transform representation for images and the singular values of the image are simple and have good stability, we transform image in the contourlet domain and decompose the singular value of low frequency component. The final image hash is obtained by applying binary quantization to the singular value. Experimental result are presented to show the effectiveness of the proposed scheme.