Binary image steganography based on joint distortion measurement

Abstract

Most state-of-the-art binary image steganography methods depend on the content of the image to determine where to embed secret messages, which is capacity-limited and indicates that their distortion measurement may be not precise enough. In this paper, we propose a kind of distortion measurement that is not only based on the discrimination effects after flipping the pixels but also depends on the visual effects of flipping corresponding pixels, which is called joint distortion measurement. Instead of selecting suitable position to embed secret messages, we then employ the syndrome-trellis code to minimize the embedding distortion and get messages embedded. And experimental results have demonstrated that the proposed distortion measurement has higher performance and the steganography scheme can achieve stronger statistical security with high capacity and image quality.