Image steganography method and system based on generative steganography confrontation

A generative and steganographic technology, applied in image data processing, image data processing, neural learning methods, etc., can solve the problems of low visual quality of generated images, small steganographic capacity, low security, etc., to improve anti-detection ability, the effect of reducing the disappearance of the gradient, and improving the ability of detection

Pending Publication Date: 2021-10-22
QILU UNIV OF TECH
View PDF0 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In the existing generative steganography technology, the visual quality of the generated image is not high enough to meet the requirements of practical use; the method of adding adversarial samples to the carrier image needs to retrain each input image, which is only suitable for a small number of images ; Divide the carrier image into regions, embed secret information in some regions, add adversarial samples to some regions, only enhance part of the image, and increase the risk of being recognized by the steganalyzer
[0007] In the existing embedded steganography technology, although the embedded steganographic model based on automatic learning

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image steganography method and system based on generative steganography confrontation
  • Image steganography method and system based on generative steganography confrontation
  • Image steganography method and system based on generative steganography confrontation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0049] Such as figure 1 As shown, the existing generative confrontation network GAN is composed of a generator and a discriminator. The generator inputs noise to generate images, and the discriminator is used to distinguish real images from generated images; the discriminator and generator are trained at the same time, and the two are constantly The adversarial game is carried out in the ground, and finally the Nash equilibrium is reached, and the generator generates the result consistent with the real image. The training process of GAN is the optimization of the minimax problem; including:

[0050] (1) The discriminator D correctly judges whether the input data comes from the real data sample x or the fake data sample G(z) generated by the generator as much as possible;

[0051] (2) The generator G learns the data distribution of the real data set samples, and tries to make the performance of the fake data G(z) generated by the generator on D consistent with the performance ...

Embodiment 2

[0133] This embodiment provides an image steganography system based on generative steganographic confrontation, including:

[0134] The model construction module is configured to construct a generative steganographic confrontation model comprising a generative network, a steganographic device, a discriminative network and a steganalytic network, wherein the generative network converts the original image into a carrier image by using the same layer jump method, so The steganography device embeds the secret image into the carrier image to obtain the secret image, the discrimination network judges the authenticity of the carrier image, and the steganalysis network judges whether the secret image contains secret image information;

[0135] The confrontation training module is configured to construct the loss functions of the generation network, the discrimination network, and the steganalysis network according to the judgment results of the discrimination network and the steganalys...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses an image steganography method and system based on generative steganography confrontation, and the method comprises the steps: constructing a generative steganography confrontation model which comprises a generative network, a steganography device, a discrimination network and a steganalysis network, enabling the generative network to convert an original image into a carrier image, enabling the steganography device to embed a secret image into the carrier image, and obtaining a secret-carrying image, wherein the discrimination network judges the authenticity of the carrier image, and the steganalysis network judges whether the secret carrying image contains secret image information; respectively constructing loss functions of the generative network, the discrimination network and the steganalysis network according to judgment results of the discrimination network and the steganalysis network, and performing adversarial training on the generative network, the discrimination network and the steganalysis network according to the loss functions of the generative network, the discrimination network and the steganalysis network to obtain a trained generative steganalysis adversarial model; and obtaining a steganographic image corresponding to the original image based on the trained generative steganographic adversarial model. A carrier image with high visual quality is generated, and the security of carrier image information hiding is improved.

Description

technical field [0001] The invention relates to the technical field of image steganography, in particular to an image steganography method and system based on generative steganographic confrontation. Background technique [0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art. [0003] Information hiding is one of the important technologies to ensure information security, which can imperceptibly hide secret information into the carrier. Information hiding can not only ensure the security of data itself, but also ensure the safe transmission of data. Steganography is an important method in the field of information hiding. In order to successfully transmit secret information, the sender uses the insensitivity of the human visual system to some areas of the digital image to hide the secret information in the carrier image in an invisible way, so that It will not arouse the suspi...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06T1/00G06K9/62G06N3/04G06N3/08
CPCG06T1/0021G06N3/084G06N3/048G06N3/045G06F18/214
Inventor 马宾韩作伟徐健马睿和李健王春鹏
Owner QILU UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products