Co-segmentation result correction method based on segmentation propagation

A co-segmentation and correct technology, applied in the field of co-segmentation technology, can solve problems such as difficulty in overcoming scene interference and difficulties in co-segmentation models

Inactive Publication Date: 2014-02-26
UNIV OF ELECTRONIC SCI & TECH OF CHINA
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, constructing efficient co-segmentation models for Lupine and
At the same time, the method based on model construction is difficult to overcome the interference of the scene, such as complex background and similar background
Therefore, the results of existing segmentation methods in practical applications need to be further improved

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
  • Co-segmentation result correction method based on segmentation propagation
  • Co-segmentation result correction method based on segmentation propagation
  • Co-segmentation result correction method based on segmentation propagation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0032] Step 1: Obtain the co-segmentation result set S, S={S 1 ,...,S m}, m is the total number of segmented images; the segmentation result is a mask mask, a binary image composed of 0 and 1, and 0 and 1 are used to distinguish the foreground area from the background area;

[0033] Divide the co-segmentation results into the correct set of segmentation results and incorrectly segmented result sets no G Correctly segment the image for m segmented images The number of i=1,…,n G , n B is the wrongly segmented image I′ among the m segmented images j The number of j=1,…,n B ; The standard segmentation result of the image here is known, does not belong to the focus of discussion of the present invention for the judgment of whether the image is segmented correctly;

[0034] Step 2: Select the wrongly segmented image I′ to be corrected j ;

[0035] Step 3: Find the current wrongly segmented image I′ j The global spectrum of :

[0036] 3-1 Take the wrong segmented ima...

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 co-segmentation result correction method based on segmentation propagation. The co-segmentation result correction method comprises the steps of dividing co-segmentation results into a correct segmentation result set and a wrong segmentation result set; selecting a wrong segmentation image I' [j] to be corrected; working out a corresponding global spectrum M<G> of the present wrong segmentation image I' [j]; working out a corresponding local spectrum M<L> of the present wrong segmentation image I' [j]; obtaining a corresponding prospect spectrum M: M = [alpha] M<L> + (1 - [alpha]) M<G> of the present wrong segmentation image I' [j] based on the global spectrum and the local spectrum in a step 6; obtaining the prospect spectrum M of the present wrong segmentation image I' [j], then segmenting the prospect spectrum M, and obtaining a corrected co-segmentation result. According to the invention, the correction method is strong in adaptive ability and accurate in correction result by global propagation and local propagation.

Description

technical field [0001] The invention relates to digital image processing technology, in particular to co-segmentation technology. Background technique [0002] With the development of the network, there are a large number of digital images in the network. Using these massive images to achieve specific object detection and segmentation has become more and more concerned. [0003] Co-segmentation is the technique of segmenting a specific object from multiple images containing the same specific object with different backgrounds. In order to segment specific objects in the current image, the existing co-segmentation methods first introduce multiple images containing the same specific object, such as images obtained by search engines, and then achieve this by extracting and segmenting the common objects contained in this group of images. Segmentation of specific objects. [0004] Although the method of digital image co-segmentation has made great progress, the effect of digita...

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): G06T7/00
Inventor 李宏亮孟凡满
Owner UNIV OF ELECTRONIC SCI & TECH OF CHINA
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