Universal steganalysis method for JPEG images

Abstract

The invention provides a universal steganalysis method for JPEG (Joint Photographic Experts Group) images. The universal steganalysis method comprises a training process and a prediction process, wherein the training process comprises the following steps: extracting the features of all images in a training sample set, and training the obtained features by a classifier to obtain a model; and the prediction process comprises the following steps: extracting the features of to-be-predicted images in the same way, and under the condition of the same classifier, predicting by the model well trained in the training process to obtain a final result. The universal steganalysis method is characterized in that all the images in the training sample set and the to-be-predicted images are used as original images for extracting the features; and extraction is realized in the following way: global calibration processing is carried out on the original images to obtain globally calibrated images, the multi-direction differential Markov probability matrixes of the original images and the globally calibrated images are respectively calculated, and differential processing is carried out on calculated results to obtain features. The detection rate of the JPEG stego images is further increased by the method provided by the invention.

Description

technical field [0001] The invention relates to the field of digital image information hiding, in particular to a general steganalysis method for JEPG images. Background technique [0002] In recent years, the continuous development of information technology and the popularization of the network have provided people with a new open communication environment, and the development of digital multimedia technology has also provided rich multimedia information for the network. But at the same time, there are also many hidden information security problems, such as hacker attacks, private information leakage, dissemination of bad information, digital media piracy and infringement, etc. are common occurrences. Information hiding technology is an important technology for confidential communication and information security. Both steganography technology and steganalysis technology are important branches of information hiding technology. People can easily use digital images as carriers...

AI Technical Summary

Problems solved by technology

[0006] The above methods can achieve effective detection of JPEG steganographic images to a certain extent, but there are still the following problems: 1) The existing steganalysis methods only consider the transition probability matrix in each direction when calculating the transition probability matrix It is only related to the difference matrix in this direction, and has nothing to do with the difference matrix in other directions. Such an assumption is unreasonable; 2) The existing steganalysis methods do not consider the characteristics of the sign correlation of frequency domain coefficients.

[0007] Based on the above problems, there are two problems in the steganalysis algorithm: 1. The detection results of steganographic images with low embedding rate are not ideal, and the detection rate of steganographic images with high embedding rate has a significant downward trend; 2. The general type is not strong enough, It is impossible to achieve an ideal detection rate for various steganographic algorithms

Images