Unlock instant, AI-driven research and patent intelligence for your innovation.

A Color Similarity-Based Edge Removal Method for Digital Image Over-segmentation

A digital image and similarity technology, applied in image analysis, image enhancement, image data processing, etc., can solve the problems of large and small interference of feature differences, inaccurate merging results, etc., and achieve the effect of reducing over-segmentation

Inactive Publication Date: 2018-11-09
EAST CHINA UNIV OF SCI & TECH
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Both of them consider the feature difference of the entire adjacent area, so the merged result is not accurate, and the size of the feature difference is easy to interfere in the larger area.

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
  • A Color Similarity-Based Edge Removal Method for Digital Image Over-segmentation
  • A Color Similarity-Based Edge Removal Method for Digital Image Over-segmentation
  • A Color Similarity-Based Edge Removal Method for Digital Image Over-segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] The present invention will be described in further detail below in conjunction with the accompanying drawings.

[0022] The flow process of the method for merging over-segmented regions with threshold value "line elimination" according to the present invention is as follows: figure 1 as shown, figure 1 Consists of 5 units.

[0023] Unit 100 is a unit for segmenting an image by JSEG.

[0024] Unit 101 is a unit for constructing a region adjacency graph. The region obtained by the JSEG segmented image obtained by the processing unit 100 is expressed as a graph structure, such as figure 2 As shown, all the divided regions are regarded as the vertices of the graph, the adjacency relationship of the region is regarded as the edge of the graph, and the feature difference between adjacent regions is defined as the weight.

[0025] Unit 102 is the threshold feature difference "three-point line elimination" algorithm, that is, as image 3 As shown, first obtain the informa...

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 digital image over-segmentation edge removing method based on color similarity, and mainly aims at eliminating over-segmentation area edges caused by a JSEG segmentation algorithm by virtue of color similarity. The digital image over-segmentation edge removing method comprises the following steps: firstly, performing statistic analysis on segmentation line information of JSEG segmentation result; secondly, acquiring primary line elimination result by using an edge removing algorithm of 'three-point line elimination;' finally, acquiring final line elimination result by using the edge removing algorithm of 'three-point line elimination.' The experiment shows that by adopting the method, an over-segmentation edge can be removed so as to improve the segmentation precision of a digital image, and the digital image segmentation result is greatly improved.

Description

technical field [0001] The invention mainly relates to digital image processing technology, in particular to a method for removing over-segmented edges of digital images based on color similarity. Background technique [0002] The JSEG image segmentation algorithm is a classic image segmentation method based on image color, and its advantage is that it considers both color information and spatial information. It fully considers the local information of the image and can obtain more accurate segmentation boundaries, but the JSEG algorithm is prone to "over-segmentation". Generally, the image preprocessing before segmentation, that is, image filtering, is combined with the segmented area to reduce "over-segmentation". At present, the commonly used merging methods are mainly threshold constraints and methods based on graph theory. The methods in graph theory mainly include region adjacency graph (RAG) and nearest neighbor graph (NNG). Both of them consider the feature differe...

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): G06T7/12G06T7/136
CPCG06T2207/10004G06T2207/10024
Inventor 张静高永伟袁玉波陈志华应方立江雷
Owner EAST CHINA UNIV OF SCI & TECH