A method and system for airspace image steganography based on generative adversarial networks

Abstract

The invention discloses a steganographic method for spatial images based on generative confrontation networks, which converts carrier images into probability maps by using U-shaped structure generative networks, and then uses a hyperbolic tangent encoding module to encode the probability maps to generate tamper point maps , and add the carrier image and the tampered point map to generate a secret image; then use the steganalysis network to distinguish the carrier image and the secret image, and feed back the classification result to the generation network in the form of an error; finally, the trained The generative network and encoding module of the algorithm are combined together as the final spatial image steganography model, which inputs the carrier image to the whole model and outputs the secret image. The invention also discloses a space domain image steganography system based on generating confrontation network, which includes a generating network module, an encoding module and an image steganographic module. The airspace image steganography method based on the generative confrontation network proposed by the present invention has obvious improvement in security and is simple in design.

Description

technical field [0001] The present invention relates to the field of information steganography, and more specifically, to a method and system for spatial image steganography based on generative adversarial networks. Background technique [0002] In the covert communication technology, there are two commonly used methods, one is cryptography, the sender encrypts the secret information through encryption technology, and the information is transmitted in the form of ciphertext, and the receiver uses the decryption algorithm and key to transmit the information after receiving the information. Interpreting ciphertext into plaintext, the attack on cryptography is called deciphering; the other is steganography, where the sender hides the secret information in the carrier and spreads it through open channels. After the receiver receives the carrier containing the secret information , to extract the secret information from the carrier. The opposite of steganography is steganalysis, w...

AI Technical Summary

Problems solved by technology

Since image steganography and steganalysis are two aspects of mutual game and mutual confrontation, the generative adversarial network in deep learning provides new ideas for image steganography and steganalysis, but considering the security and self-sufficiency of steganography Adaptability, the field of spatial image steganography still does not have a good solution

Images