Research On Information Hiding Based On Grey Prediction Model

The emergence and development of information hiding technology provides an effective technical means for the copyright of digital product, covert communication and authentication. the exploration and improvement of the information hiding technology is an important research topic in the field of information security. This thesis has successfully applied the grey system theory to the information hiding technology based on image. The main work and innovation are as follows:1. The thesis studies the principle and model of information hiding technology and information hiding technology based on grey correlation analysis and Contourlet transformation. So the thesis firstly introduces the grey correlation analysis and describes the application of grey correlation analysis in image edge detection, that makes a thorough study of the theory of Contourlet transformation. The study for the theoretical knowledge has laid a solid foundation for the further research work of the thesis.2. The thesis has put forward an improved algorithm based on Tian’s after the thorough study on the spatial algorithm. Tian’s algorithm embeds the secret information by modifying the values of adjacent pixels, the modified pixels may produce overflow.As to the kind of pixels, Tian gives up them directly, which decreases the capacity of embedded information. The thesis proposes an improvement algorithm based on Tian’s.Before embedding, we adjust the carrier image by using grey correlation analysis.Firstly, divide the carrier image into blocks. Secondly, calculate the correlation value in the horizontal direction, vertical direction and diagonal direction, then adjust the largest correlation value to the horizontal direction. Finally, embed the secret information. The experiment results show that the algorithm not only has good imperceptibility, but also has larger embedding capacity.3. The thesis has put forward an algorithm based on grey correlation degree and Contourlet transformation after the thorough study on the transform domain algorithm and Contourlet transformation. The algorithm gets the parent coefficient and child coefficient through Contourlet transform. And choose the region, where the parent coefficient and child coefficient have similar energy, to embed the information. The algorithm not only can avoid producing image distortion when we only embed thesecret information into the region with high energy region, but also can improve the robustness of the algorithm. In order to improve the security of the algorithm, conduct the image sharing to the secret information using Chinese remainder theorem before embedding. The experiment results after carrying some common attacks on the algorithm show that the algorithm has visually imperceptible, and the extracted secret information has a certain degree of recognition when the image with secret information is under serious attack, so the algorithm has good robustness, and can resist the various attacks.