An efficient information hiding method based on motion vector space encoding for HEVC

As the newest video coding standard, high efficiency video coding (HEVC) has great potential as a new information hiding carrier. This paper proposes an efficient information hiding method based on motion vector space encoding for HEVC encoding process. In this method, the mapping relationship between motion vector set and the points in the motion vector space is defined. The motion vector components from the N/2 prediction units (PUs) with smallest size in a coding tree unit (CTU) are selected as the secret information carriers. Each N secret bits are converted to a 2N + 1-ary number. By modifying at most one element in the set of N motion vector components, the mapping value of the set in the motion vector space can be equal to the 2N + 1-ary number. In this way, information hiding is realized. Since at most one element is changed and the N/2 PUs with smallest size are selected, this method contributes to excellent transparency of steganography and anti-steganalysis performance with high embedding efficiency. To the best of our knowledge, this is the first information hiding method based on motion vector for HEVC. Experimental results verify that the proposed method is practicable and has better performance than two typical embedding rules of information hiding based on motion vector.