Research Of Image Steganalysis In Spatial Domain Of Images

With the rapid development of network communication technology and the increasing popularity of digital multimedia data, messages can be easily transmitted through the network. At the same time, an increasingly important data security issues are revealed. Especially, the rapid development and wide application of steganography lead to the transmission of confidential information becoming more covert. Now, Military agencies, government departments, financial institutions and other key sectors involved in the people's livelihood take advantage of the data embedding technique to achieve secure transfer of confidential documents. However, this technique can easily be used by malicious individuals and groups for a variety of illegal purposes, thus posing serious threat to the national security, social stability as well as the personal privacy. Steganalysis attempts to break the hidden information, and hence, it is an important mean of combating the illegal information transmission. Thus, research on the detection of hidden information is significantly important to ensure the public safety, national security, social stability, and the normal economic development.In this paper, we take the digital image as the research object, and the research includes information detection methods with the steganography of replacing least significant bit (LSB) in space domain image, rate estimation of hidden information. The main contributions of this paper are listed as follows.Based on two LSB steganography in space domain, we proposed the least square method based on weighted stego image. In this detection method, we first construct a weighted image and use pixel estimator to get an estimated image which approximated to the cover image, then construct an equation by the two estimated images, we can get the embedded message rate through the equation. Experimental results show that this method outperforms the current detection methods.A steganalysis method based on the finited state machine is proposed. In this method, we take the pixel pair as different transform states, and then analyze the relation of transformed states in order to construct the finited state machine. Message rate equation can be constructed through the relation of the state transformed. Experimental results show that the proposed detection method improve the detection rate and save a lot of time in computing the states.An extended optimization method is proposed to detection the steganography of multiple LSB replacement. In this detection method, first we introduced the symmetry of image pixel, and then we extended the optimization detection method in order to detect the embedding rate of multiple LSB replacement,finally the experimental results show that our method can ave the computational time,and make the detection more reliable．...