Implicit Write And Spatial Image Detection Technology Based On LSB

With the rapid development of computer and Internet technology, network and digital technology has entered a new era. Steganography and steganalysis detection has been gradually treasured by governments, military institutions and various research organizations. Steganography can be used to transmit confidential documents, but it can also easily be used by criminals for illegal purposes, such as the spread of illegal information. Therefore, researches of steganographic testing technology are of great significance to national information security.LSB steganography algorithm has been widely used for its simple implementation and large capacity. With digital images as the research object, this article makes an in-depth research on the least significant bit (LSB) steganographic methods and related steganalysis method. The main research results are as follows:Firstly, we use Arnold scrambling to the secret image to make it invisible, then by adding an extra bit for exclusive or operation; we can change the statistical characteristics of the original method. This method increase the detection of resistance while the encrypt capacity are cut down.Aiming at sequential embedding and random embedding method of LSB algorithm, this paper implements the chi-square detection algorithm and RS detection algorithm respectively. With these algorithms we can get precise information embedding rate of an encrypted image.Based on the evaluation of embedding capacity of existing airspace least two bits encrypt method, this paper implements a minimum mean square error detection algorithm based on weight. The detection algorithm firstly constructs a weighted image and re-constructs a similar image using the pixel estimation method, then uses this two images to construct minimum MSE equation to get the length of encrypt information.Based on the algorithm of evaluation on embedding rate of weighted encrypted images, this paper makes improvement on the Filter mask, so that the new algorithm can get the exact embedding rate.