Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Convolutional neural network-based digital image steganalysis method

A convolutional neural network and digital image technology, applied in the field of digital image steganalysis based on convolutional neural network, can solve problems such as low analysis performance and complex effective feature design, and achieve high accuracy

Inactive Publication Date: 2017-11-07
SUN YAT SEN UNIV
View PDF2 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The present invention provides a digital image steganalysis method based on a convolutional neural network in order to solve the technical defects of low analysis performance or complex effective feature design in the steganalysis method provided by the above prior art.

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
  • Convolutional neural network-based digital image steganalysis method
  • Convolutional neural network-based digital image steganalysis method
  • Convolutional neural network-based digital image steganalysis method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0021] Such as image 3 As shown, the convolutional neural network consists of multiple convolutional layers connected in series, and the input of the latter layer is the output of the previous layer. Each convolutional layer contains the following three basic operations: convolution, nonlinear activation, and pooling.

[0022] The process of convolution operation is as follows figure 1 shown. The input data is convolved with a convolution kernel of a predefined size, and the corresponding convolution feature map can be obtained. One convolution kernel corresponds to one feature map. When multiple convolution kernels are used, multiple feature maps will be output.

[0023] Non-linear activation is to transform the feature map obtained by convolution point-by-point using a non-linear activation function. A commonly used activation function is Rectified Linear Unit (ReLU), which is defined as:

[0024] f(x)=max(0,x)

[0025] That is, keep all signals greater than 0, and se...

Embodiment 2

[0044] The convolutional neural network used in the present invention does not need complex artificial feature design, and can complete feature extraction and classification within a set of frameworks, so that each step can be optimized at the same time, greatly reducing the difficulty of steganalysis algorithm design . In order to verify the performance of the proposed algorithm, the present invention implements the proposed algorithm based on the deep learning framework Caffe. The proposed network is referred to as TLU-CNN in the following. All experiments are carried out on image databases BOSSBase and BOWS2, which are commonly used in steganalysis. In order to be able to compute with GPU, the training and testing image sizes are scaled from 512×512 to 256×256.

[0045] In this embodiment, the current leading digital image steganalysis algorithm——SRM is selected as a comparison, and the experimental results obtained on the three currently safest three adaptive steganograp...

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 present invention designs a convolutional neural network-based digital image steganalysis method. The convolutional neural network-based digital image steganalysis method comprises the following steps of S1 constructing a convolutional neural network formed by connecting multiple convolutional layers in series; S2 for the first convolutional layer, adopting a high-pass filter to initialize a convolution kernel, and then adopting a truncation linear unit activation function as the activation function of the convolutional layer; S3 inputting a digital image in the convolutional neural network, and using the convolutional neural network to output a result of the digital image whether or not after the steganography.

Description

technical field [0001] The present invention relates to the technical field of digital image steganalysis, and more specifically, to a digital image steganalysis method based on a convolutional neural network. Background technique [0002] Digital image steganalysis is a very important research direction in the field of information security. The so-called steganalysis refers to the technology of detecting images embedded with secret information by using a certain method. Since the design of the steganographic algorithm focuses on the impact of the embedded information on the original image, the secret image is very close to the original image both visually and statistically. In order to prevent steganography from being used in illegal ways, effective methods are needed to detect this kind of covert communication behavior. Efficient steganalysis methods can extract highly discriminative statistical features from images, so as to capture the traces left when steganography mo...

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): G06T1/00G06N3/04G06N3/08
CPCG06N3/08G06T1/0021G06N3/048G06N3/045
Inventor 倪江群叶健
Owner SUN YAT SEN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products