Capacity Analysis Of Information Hiding

In the process of information hiding, the original information is modified by adding secret message and transferred to the receiver. The embedding should be transparent (i.e., the modified data sequences or "stegotext" should be similar to the covertext). In order to meet the requirement of transparency of information hiding process and maximize the amount of secret message, the concept of capacity of information hiding is brought up. The capacity of information hiding refers to the amount of message which is embedded into the covertext. It is affected by many factors such as the data model of original image, distortion constraint function and the attacking channel, etc. Research indicates that for a certain attack distortion D₂, an optimal result of capacity is achieved by making use of Gaussian model for the original image and Minimum mean square error for distortion and using the Gaussian test channel . Meanwhile the game model and EP model are investigated and how to compute the capacity of information hiding for each model is illustrated later. At last an improvement for previous models is put forward which exploits the principle of correspondence between subjectivity and objectivity for the computation of the capacity.