A GA optimization approach to HS based multiple reversible data hiding

As a typical reversible data hiding scheme, the performance of histogram shifting (HS) technique is dependent on the selected side information, i.e., peak and zero bins. Due to the massive solution space and burden in distortion computation, the conventional HS based schemes utilize some empirical criterions to determine those side information. In this paper, the rate and distortion model and the associated fast algorithm of distortion computation is first developed, the HS based multiple embedding is then formulated as the rate and distortion optimization problem. An intelligence optimization algorithm, i.e., genetic algorithm (GA) is employed to automatically search the globally optimal zero and peak bins. For a given secret data, the proposed scheme could not only adaptively determine the optimal number of peak and zero bin pairs but also their corresponding values for HS based multiple reversible embedding. Compared with previous approaches, experimental results demonstrate the superiority of the proposed scheme in the terms of embedding capacity and stego-image quality.