Steganalysis method and device based on least significant bit matching

Abstract

The invention relates to a steganalysis method and device based on least significant bit matching, which divides an image sequence into several disjoint blocks, and each block satisfies a multivariate Gaussian distribution to obtain the first expression of the image sequence; according to the least significant bit The matching algorithm is used to determine the second expression of the image sequence after embedding the hidden information; according to the first expression and the second expression, the likelihood ratio expression is determined; according to the Taylor expansion algorithm, the likelihood ratio expression is processed to determine The detection subexpression of each block; divide the observation image into disjoint blocks, and determine the detection subvalue of each block in the observation image according to the detection subexpression; analyze the smoothness of the observation image to determine the observation image The smoothness value of each block in ; according to the preset smoothness threshold, sum the detector values ​​of the blocks satisfying the preset smoothness threshold to obtain the judgment detector of the observed image; according to the judgment detector of the observed image, determine Observe whether the image is a steganographic image.

Description

technical field [0001] The invention relates to the field of information hiding analysis, in particular to a steganalysis method and device based on least significant bit matching. Background technique [0002] Information Hiding is a research hotspot in the field of information security, and has been widely used in military, political, commercial and other fields. Information hiding includes digital watermarking (Digital Watermarking) and image steganography (Steganography) two branches. These two branches have many similarities in technical realization, but each has its own characteristics, and its purpose is very different. The purpose of digital watermarking technology is to protect the copyright of multimedia data, so that more emphasis is placed on the robustness of secret information; while the purpose of image steganography is to deliver information secretly and safely, because it is necessary to avoid causing the second party as much as possible. Suspicion by the ...

AI Technical Summary

Problems solved by technology

[0003] For image steganography, the existing technology proposes a steganalysis algorithm based on likelihood ratio test for least significant bit matching algorithm detection. A serious problem of this algorithm is that it assumes that the carrier image obeys independent Gaussian distribution, this assumption obviously does not conform to the natural image laws in the real world, because there are obvious correlations and redundancy between a large number of pixels in natural images, and the independent Gaussian distribution obviously ignores this redundancy In addition, the algorithm sums the detection sub-values ​​of all pixels without weighting when considering the image detection sub-values, which affects the accurate detection of steganographic information

Images