Image steganography method and system for automatically learning distortion based on GAN network structure

Abstract

The invention discloses an image steganography method and system for automatically learning distortion based on a GAN network structure, and the method comprises the steps: inputting an enhanced carrier image into a preset improved U-Net network, sequentially generating an initial pixel change probability graph and an initial secret-containing image, inputting the initial secret-containing image into a discrimination network, and obtaining a discrimination result, calculating the loss of the discrimination network according to the discrimination result, calculating a total loss function of the generative network according to the discrimination result and the current steganography capacity, optimizing the generative adversarial network by taking minimization of the loss function as a target, when the loss is reduced and kept stable, considering that the training is ended, extracting the generative network from the generative adversarial network after the training is ended, and inputting a to-be-transmitted original image into the generative network to obtain a pixel change probability, calculating embedded distortion corresponding to the pixel change probability, and coding the secret information and the to-be-transmitted original image by adopting a syndrome matrix coding technology according to the embedded distortion to obtain a steganographic image corresponding to the original image.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image steganography method and system for automatically learning distortion based on a GAN network structure. Background technique [0002] Image steganography is a technology that hides secret information in carrier images for covert communication. This technology plays an important role in many aspects such as information security and data communication, and is of great significance to ensuring data security and maintaining social stability. However, there are great challenges in image steganography: the criterion for defining the steganographic distortion cost function is the heuristic principle, and the process of designing the distortion cost function depends on human experience knowledge, and lacks the verification interaction of statistical detection. [0003] In recent years, the research and development of deep learning has entered an explosive period, and m...

AI Technical Summary

Problems solved by technology

However, there are still some problems in these methods, such as unstable training, poor image quality affecting the security of steganography, etc.

Images