Optimal quantization-based steganography for adaptive hiding

In this paper, we proposed a general framework for designing a good algorithm based on the optimal quantization-based steganography (QBS). After the description of the concepts of QBS, we present an implement of simple QBS(SQBS) on the pair value difference domain. In our implement, we adopt the weighted sum of the distance between the cover and stego signals to measure the distortion, and construct an optimization problem with the synthesized index combined by the capacity and distortion as the goal, and the capacity and distortion as the restriction. Through the dynamic programming, the optimal Q-book is obtained to perform adaptive data hiding. Our experiments show the proposed method can achieve good performance both on the capacity and distortion.