An Image Steganography Based On Wavelet Domain

Steganography is the means of concealed communication, its purpose is toembed secret information into cover objects, so that the presence of secret informationcan not be detected. The opposite side is steganalysis which aims at detecting thepresence of secret information. The steganography on wavelet domain based onhuman visual system have the characteristics of large embedding capacity and goodimperceptibility, therefore it occupies an important position in the research ofsteganography.So the paper focus on the analysis and realization of two typical steganographicalgorithms on wavelet domain based on human visual system, and find that thosesteganographic algorithms result in abnormity of wavelet coefficient histogram, thenpropose the new steganographic algorithm. The main important work of this paper：1. Deeply analysis image steganography, steganalysis, human visual system, andthe basic theory of wavelet transform.2. Detailed introduce the steganographic algorithm based on Bit-PlaneComplexity Segmentation(BPCS) and Integer Wavelet Transforms(IWT), and thesteganographic algorithm on wavelet domain based on Human Visual System(HVS)against statistical analysis, then point out theory analysis and experiment simulation.The results show that the steganography on wavelet domain based on human visualsystem have the characteristics of large embedding capacity and good imperceptibility,but those steganographic algorithms result in abnormity of wavelet coefficienthistogram.3. A novel steganalysis algorithm about Wavelet Coefficients StatisticalFeatures(WCSF) is presented. By analysing the wavelet coefficients histogram ofcover-images and stego-images, we find that they have significant difference, anddelimit the WCSF. Then, by calculating the WCSF of images, we judge the presenceof secret information. Finally, we propose the WCSF steganalysis algorithm. Theoryanalysis and experimental results show that this steganalysis algorithm can announcethe presence of hidden information.