Steganalysis of Palette Images: Attack Optimal Parity Assignment Algorithm

Steganography is the art of communicating a message by embedding it into multimedia data without drawing any suspicions. Optimal parity assignment (OPA) algorithm is one of palette image steganographic methods. But there is still no effective steganalytic method that can attack it reliably. In this paper, we present a statistical steganalysis for detecting the secret messages embedded by using OPA algorithm. Based on the analysis of RTS (replacement-transfer structure) of index replacement in GIF-based steganography, we explore the singular statistic of core elements in OPA stego-images via Convergent Continuous Replacement (CCR) operation. On the other hand, we can estimate the statistic of cover image from stego-image by a special filtering operation. With the estimation of cover image statistic and the singular statistic of the stego-image, the length of secret messages embedded in GIF images with OPA algorithm can be estimated. The experimental results indicate that our algorithm can estimate the length of hidden messages accurately.