Adaptive image steganalysis method and system based on deep convolutional neural network

Abstract

The invention relates to the technical field of image steganalysis, and discloses an adaptive image steganalysis method and system based on a deep convolutional neural network. The system comprises amodel training module, a feature preprocessing module and a steganalysis detection module. The method comprises the following steps: carrying out information embedding on a gray image through employing an adaptive image steganalysis algorithm, and manufacturing a steganalysis image data set; processing the gray image of the data set through a high-pass filtering layer and two cascaded 3 * 3 convolutional layers to obtain a residual image, and training the residual image on a deep convolutional neural network model; then continuously adjusting network parameters and structures, continuing training, and finally screening an optimal deep convolutional neural network model; selecting a gray image to be detected, and then performing high-frequency feature extraction on the image through the high-pass filtering layer and the two cascaded 3 * 3 convolutional layers; and inputting the extracted features into the optimal deep convolutional neural network model, carrying out steganography imagedetection, and outputting a detection result. The advantages of the deep convolutional neural network are utilized, so that the efficiency and accuracy of adaptive steganography image detection can beeffectively improved.

Description

technical field [0001] The invention relates to the technical field of digital image steganography, in particular to an adaptive image steganography analysis method and system based on a deep convolutional neural network. Background technique [0002] In recent years, with the development of steganographic technology, the emerging adaptive steganography has gradually become a popular research direction in the field of image steganography and steganalysis. This algorithm combines the structural characteristics of the image itself to adaptively select images that are relatively difficult to detect. , Insensitive areas for message embedding. The current mainstream steganalysis methods mainly rely on artificially designed features, which require a lot of time and energy, and the current results cannot meet the actual needs. Deep learning does not require artificial design features. It is a new solution. By constructing a learning model composed of multi-layer linear and nonline...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to propose an adaptive image steganalysis method based on a deep convolutional neural network for the deficiencies of the existing steganalysis methods

Images