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

A method and system for image superpixel segmentation

A superpixel segmentation and image technology, applied in the field of image processing, can solve the problem that the target area segmentation cannot achieve the desired effect, affect the accuracy of target segmentation, and cannot adjust the calculation strategy, and achieves good edge recall rate and good edge fit. , the effect of preventing edge loss

Active Publication Date: 2021-03-09
北京中科晶上超媒体信息技术有限公司
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the traditional SLIC algorithm only considers the color features and distance position information between pixels, and cannot adaptively adjust the calculation strategy according to the characteristics of different images, resulting in the segmentation of the target area in some images with complex colors. The ideal effect cannot be achieved, and overflow or incomplete segmentation occurs, which affects the accuracy of subsequent target segmentation and extraction.

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 method and system for image superpixel segmentation
  • A method and system for image superpixel segmentation
  • A method and system for image superpixel segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0046] After testing the traditional SLIC algorithm, the inventor found that the main reason why it cannot achieve the desired effect is that the color distance and the space distance are set to have the same weight when performing image segmentation. To illustrate this point, the traditional SLIC algorithm will be introduced below. Operations based on this algorithm include:

[0047] a) Initialize the seed points (that is, the cluster centers), and distribute the seed points evenly in the input image according to the number of superpixels set by initialization. Assuming that the input image has N pixels, the number of initial superpixels is K, then the size of each superpixel is N / K, and the distance between adjacent cluster centers is approximately S=sqrt(N / K);

[0048] b) Reselect the seed point in the small neighborhood around the seed point, that is, take the position with the smallest gradient to avoid the seed point falling on the boundary of the contour;

[0049] c) ...

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 provides a method for superpixel segmentation of an image, the image contains one or more image blocks, the method includes: 1) determining the richness of color information contained in each image block; 2) according to The richness of the color information contained in the image block determines the coefficient α for the color distance and the coefficient β for the spatial distance set for the image block in the SLIC algorithm. This way makes it possible to adaptively adjust the influence of color and pixel position on the segmentation results according to the image characteristics when implementing the SLIC algorithm. Through the test ratio, the method of the present invention can obtain better edge recall rate and under-segmentation error rate at the cost of a very small increase in running time, that is, the method of the present invention can better prevent edge loss and obtain more Good edge fit.

Description

technical field [0001] The present invention relates to image processing, in particular to image area division. Background technique [0002] With the development of science and technology, more and more applications require computers to replace human eyes to complete the operations of identifying, tracking, and long-term analysis of target areas in video frames. Different from the image background area, the target area is the hotspot that human eyes focus on. How to let the computer determine the target area is the key to this technology. [0003] Superpixel segmentation is the preprocessing to achieve operations such as object region segmentation, and has been widely used in computer vision applications such as image segmentation, pose estimation, and object recognition. It was first proposed in the article "Learning a Classification Model for Segmentation" published by Ren.X et al. in ICCV in 2003. Irregular pixel blocks of visual significance are used as a superpixel, ...

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): G06T7/11G06T7/90
CPCG06T7/11G06T7/90
Inventor 刘畅赵潇高明晋周一青石晶林
Owner 北京中科晶上超媒体信息技术有限公司
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