Research On Steganography For Document Images

Information hiding technology is a hot area in the research of information security, which has wide applications in many fields such as copyright protection of digital media, hidden communication. Along with the development of digital image collecting, processing, transmission and application technology, it's more and more prevalent to change varieties of papers containing document information to digital images, ie, document images. The information hiding technology using document images as carrier is of obvious value both in theory and application research. But the research for steganography techniques of document images is less.This dissertation works on steganography techniques in document images based on the analysis of gray document images. The main work and contribution of this dissertation are summarized as followings:1,The concept of the document image is explained. The differences between the document image and the ordinary image bearing continuous hue for the data distribution features in space domain as well as in transform domain are compared. The performances of the document image steganography are introduced.2,A information hiding algorithm with large capacity for gray document images is proposed. In consideration of data distribution and vision effect in bit planes of the gray document image, the secret information is embedded in the first and second least significant bit plane when the gray value of the pixel is equal to or larger than 240. If the gray value of the pixel is less than 240, the secret information is embedded in the least four significant bit planes. The algorithm using key to select position for embedding information. This algorithm has improved the hiding capacity while achieves security. The hidden information is blindly extracted. The experimental results show that this method is simple and has big capacity.3,Introducing the PVD algorithm to document image steganographic technique, a PVD algorithm based on gray document image is proposed. According to the gray value distribution features of document images, this algorithm adopts different ranges for background pixels and other pixels to embed information, which can improve the embedding capacity and has some resistance to RS detection. The embedded secret message can be extracted form the resulting stego-image without referencing the original cover image. Experimental results show the feasibility of the proposed method.Finally, the research work for the dissertation is summarized, and the promising directions for future research is identified.