An optimized algorithm based on generalized difference expansion method used for HEVC reversible video information hiding

Abstract

HEVC, the latest video coding standard, is jointly developed by ISO/IEC and ITU-T. Based on HEVC, an optimized generalized difference expansion method for reversible video information hiding is realized. Firstly, 2N+1 values of the spatial point are obtained by constructing the N-dimensional motion vector space. Then the optimized generalized difference extended algorithm is used to embed the secret information into the value of the spatial point. Finally, without using the location map, this method can help to embed 2N + 1 bit secret information into the motion vector N-tuple C, and can recover the carrier video with little distortion during the extracting process. The experimental results show that the false noise ratio of the video image is more than 32db after the steganography process, and the bit rate is no more than 0.77%. These characteristics mean that the implement of the method will not distort the video and has good visual invisibility. During the extracting process, the embedded secret information can be extracted accurately. What's more, the original video can be restored without distortion. Therefore, this method is very useful for realizing the secret information hidden and restoring the carrier video without distortion.