Texture distribution weak hypothesis and regularization strategy-based natural image matting method

A natural image matting and texture technology, applied in image analysis, image data processing, instruments, etc., can solve problems such as lack of accurate analytical solutions and under-constrained

Inactive Publication Date: 2016-01-06
WUHAN UNIV
View PDF2 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the matting problem is a typical under-constrained problem, and there is no precise analytical solution. It has always been a challenging problem in the field of computer image processing.

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
  • Texture distribution weak hypothesis and regularization strategy-based natural image matting method
  • Texture distribution weak hypothesis and regularization strategy-based natural image matting method
  • Texture distribution weak hypothesis and regularization strategy-based natural image matting method

Examples

Experimental program
Comparison scheme
Effect test

test Embodiment w

[0051] Test case collection of test graphs available at www.alphamatting.com.

[0052] Sampling processing stage:

[0053] Step 1: Construct trimap, divide the image into three parts, known foreground area, known background area and unknown area to be solved;

[0054] Step 2: Take a point p in the unknown area to be solved, assuming that the coordinates are (x, y); collect samples in the area centered on point p. Collect foreground samples and background samples within a radius r around p;

[0055] Step 3: Use the OM clustering method to cluster the samples, and obtain the weighted mean value of the clustered result prospects and the weighted mean of the background Foreground covariance matrix Σ F and the background covariance matrix Σ B ;

[0056] Color processing stage:

[0057]Step 4: Establish the mathematical relationship model P(F,B,α|C) of color and opacity through the Bayesian framework, and then use the maximum posterior probability MaximumAPosteriori——MAP to...

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 provides a texture distribution weak hypothesis and regularization strategy-based natural image matting method. An improvement is made aiming at the defects of the Bayesian matting algorithm. Firstly, aiming at the problem that an original method is too strong in hypothesis, the hypothesis is weakened; and a calculation measure and a correction coefficient for carrying out correction on a Gaussian distribution variance are defined and designed on the basis of a histogram and the bhattacharyya distance, so that an adaptive variance Gaussian distribution model of considering texture complexity is provided; and complicated texture distribution of natural images is effectively coped with. Secondly, aiming at a calculation model of the original method, a regularization strategy-based matting algorithm model is provided on the basis that an ideal corrective color estimation optimization model and a solving problem thereof are provided, namely a data constraint term and a penalty term are added to a basic model through an augmented lagrangian multiplier method; and a solving model, in a regular form, of an ideal optimization model is obtained.

Description

technical field [0001] The invention relates to the field of image matting, in particular to a natural image matting method based on a weak assumption of texture distribution and a regularization strategy. Background technique [0002] Image matting is a hot issue in image processing, and matting is to separate a part of the image from the whole image. The separated part is called the foreground, and the remaining part is called the background. Cutout is widely used in digital image editing, film and television special effects production, virtual reality and other fields. However, the matting problem is a typical under-constrained problem without an accurate analytical solution, and has always been a challenging problem in the field of computer image processing. [0003] There are two main types of current mainstream matting algorithms: sampling-based methods and propagation-based methods. Sampling-based methods use neighborhood pixel information, while propagation-based ...

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 Applications(China)
IPC IPC(8): G06T7/00G06T7/40
Inventor 何发智陈晓潘一腾张德军
Owner WUHAN UNIV
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap