Digital Video Steganalysis Exploiting Statistical Visibility in the Temporal Domain

In this paper, we present effective steganalysis techniques for digital video sequences based on interframe collusion that exploits the temporal statistical visibility of a hidden message. Steganalysis is the process of detecting, with high probability, the presence of covert data in multimedia. Present image steganalysis algorithms when applied directly to video sequences on a frame-by-frame basis are suboptimal; we present methods that overcome this limitation by using redundant information present in the temporal domain to detect covert messages embedded via spread-spectrum steganography. Our performance gains are achieved by exploiting the collusion attack that has recently been studied in the field of digital video watermarking and pattern recognition tools. Through analysis and simulations, we evaluate the effectiveness of the video steganalysis based on linear collusion approaches. The proposed steganalysis methods are successful in detecting hidden watermarks bearing low energy with high accuracy. The simulation results also show the improved performance of the proposed temporal-based methods over purely spatial methods