High-capacity data protection method based on GAN augmented image foreground object

Abstract

The invention discloses a high-capacity data protection method based on a GAN augmented image foreground object. By applying various GANs to information hiding, the high-capacity data protection method obtains an embedded distortion probability graph, calculates an unsuitable steganography area according to the embedded distortion probability graph, generates a foreground object in the unsuitablesteganography area through a fixed area object generation and information embedding module, and then hides the secret information in the generated graph. The high-capacity data protection method mainly comprises the following steps: (1) inputting an original image into a steganography probability graph generator; (2) calculating the steganography probability graph; (3) calculating the unsuitable steganography area; (4) adding a texture object into the original image; and (5) carrying out an information hiding algorithm to embed secret information. According to the high-capacity data protectionmethod, the unsuitable hidden area is calculated for the image in advance, so that the directional hiding capability of the unsuitable hidden area of the original image in information hiding is improved; and provided is a high-capacity data protection method which increases hiding capacity and has high security.

Description

technical field [0001] The invention relates to a high-capacity data protection method, in particular to a high-capacity data protection method based on GAN amplified image foreground objects. Background technique [0002] With the advent of big data and the rapid development of science and technology, data security and privacy protection have become the focus of the country and individuals. With the development of the information age, information hiding technology in the field of information security has been widely proposed in order to protect secret information more covertly and safely. [0003] Information hiding refers to using the redundancy of digital carriers (including text, images, audio, video, etc.) to hide secret information in them to obtain encrypted carriers. Since the secret information is imperceptible after embedding, the secret carrier can be transmitted to the receiving party without arousing the suspicion of the monitoring party. The receiver extracts...

AI Technical Summary

Problems solved by technology

However, this method has the following problems: 1) Although GAN is used to generate foreground objects suitable for embedding information, the hiding ability of the original image is not judged, which may cause the generated foreground object to not be located in the inappropriate hidden area of ​​the original image; 2) it can The premise of the realization is that the original image does not contain foreground objects, only the background image

Therefore, the range of images targeted is limited, rather than arbitrary images, which leads to limitations of this method; 3) In the implementation process, it is necessary to reconstruct the background of the original image through GAN, which is computationally intensive and will reduce the visual preservation of the original image. true degree

Images