A Histogram-preserving Steganography Based on Wet Paper Coding and Graphic Matching Theory

This paper presents a novel double-layered embedding steganographic algorithm based on steganographic codes, wet paper codes and graphic matching theory, which can achieve high embedding efficiency and preserve the first-order histogram of the cover image. At first, a new 4-ary steganographic channel is constructed. Then the secret message is embedded in the LSB (Least Significant Bit) layer of this channel using binary steganographic codes and in the second LSB layer using wet paper codes, and the elements that need to be changed, are obtained. Afterwards, a graph is constructed. Then a changing manner of the cover image is found to decrease the distortion and preserve the histogram, using the matching algorithm in graph theory. The performance analysis and experimental results show that PSNR, the embedding efficiency and the ability to preserve the first-order histogram of the proposed algorithm are better than the steganographic software Steghide, when the same amount of secret message is embedded.