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

Multi-geometric attack resistant robust watermarking method

A robust watermarking and geometric attack technology, applied in image data processing, instrumentation, image data processing, etc., can solve the problems of impossible restoration of original images, irreversible shearing attacks, and difficulties in designing robust watermarking algorithms.

Inactive Publication Date: 2012-05-30
BEIJING JIAOTONG UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Due to too many parameters of this type of attack, it is very difficult to design a corresponding robust watermarking algorithm
After being cut, the original image cannot be recovered because the lost content cannot be repaired, so cutting is an irreversible attack

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
  • Multi-geometric attack resistant robust watermarking method
  • Multi-geometric attack resistant robust watermarking method
  • Multi-geometric attack resistant robust watermarking method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0047] The detailed description of the robust watermarking method against multiple geometric attacks proposed by the present invention is as follows in conjunction with the accompanying drawings.

[0048] 1. Construction of invariant feature space

[0049] The construction of the invariant feature space specifically includes the following steps:

[0050] Step 1: Compute the Directional Convolution Map

[0051] First, we compute the orientation graph G for H directions i , 1≤i≤H,G o(u, v) represents the gradient value of the pixel at the position (u, v) in the o direction. In order to keep the polarity of the pixel gray value change unchanged, set its value to zero when the value is less than zero. formalized as Here I represents the input image, o represents the direction of the gradient, (g) + The operation operator represents (a) + =max(a,0). Then use a different Gaussian kernel G for each direction map ∑ Do multiple convolutions to get convolution maps with differen...

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 a new multi-geometric attack resistant robust watermarking method. At present, most digital watermarking methods can only resist global transformation geometric attacks such as rotating, zooming, translating, other affine transformations and the like, but has poor resistance to a shearing attack and a local geometric attack (such as any distortion attack). The invention discloses a mapping algorithm from an image pixel point to multi-dimensional invariant feature space; geometric invariant spatial binary tree is established when feature space is divided; and digital watermarking information is embedded into the leaf nodes of the spatial binary tree bit by bit so as to fulfill the aim of resisting geometric attacks. By using the method, the attacks of image compression (such as JPEG compression), low-pass filtering, rotating, translating, shearing geometric transform, linear geometric transform and the like can be resisted, any geometric distortion attacks such as random jitter, global bending, high-frequency and the like can be resisted, and a large-amplitude shearing attack can be resisted. Due to the adoption of the method, the copyright of digital media can be protected effectively.

Description

technical field [0001] The invention belongs to the field of multimedia signal processing, in particular to a new digital watermark method which can resist various geometric attacks. Background technique [0002] With the rapid development of computer and Internet technology, multimedia files are widely disseminated, and the protection of their intellectual property becomes more and more important. Digital watermarking is an effective method, which embeds specific copyright information into the image as a proof of copyright owner. [0003] To play a role in protecting copyright, digital watermark should have two basic characteristics: first, transparency, that is, the image embedded with watermark is almost the same as the original image, the existence of digital watermark cannot be seen, and the original work should not be damaged as much as possible; secondly It is robustness, that is, digital watermarking should be able to resist various intentional or unintentional atta...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06T1/00
Inventor 赵耀田华伟倪蓉蓉秦伦明
Owner BEIJING JIAOTONG 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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com