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

A Robust Global Threshold Segmentation Method

A global threshold and robust technology, applied in the field of image processing, can solve problems such as unsatisfactory results and uneven cumulative probability distribution, and achieve the effect of good versatility and strong robustness.

Active Publication Date: 2019-03-12
INST OF SEMICONDUCTORS - CHINESE ACAD OF SCI
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, when the area, variance, and cumulative probability distribution of the image foreground and background are not uniform, the Otsu threshold method often achieves unsatisfactory results.

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 Robust Global Threshold Segmentation Method
  • A Robust Global Threshold Segmentation Method
  • A Robust Global Threshold Segmentation Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment example 1

[0047] Such as figure 2 , Shows an image with a large difference in gray distribution between the foreground and the background and the corresponding histogram, T o With T b And Ts are 126, 142, 85, λ=0.127, which is greater than th. It shows that the Otsu threshold will be biased towards the low gray level, which is the background in this image. Therefore, the optimal criterion adopted is p 1 (m1mg) 2 , At this time, it shows that the variance of the high-gray-level image compared with the low-gray-level image has no effect in determining the optimal threshold. After adopting the optimal criterion, the obtained threshold is equal to 85, that is, the optimal threshold is obtained, T f = T i . image 3 (a), 3(b), 3(c), 3(d) respectively indicate that the threshold is equal to T o , T b , T i The segmentation results and manual segmentation results. The result shows that the present invention successfully segmented the foreground object, but the Otsu threshold method failed.

Embodiment example 2

[0049] Such as Figure 4 (a) shows an image with poor contrast and far greater foreground variance than background variance. Figure 4 (b) is the result after the histogram equalization. It can be seen that the contour of the foreground is not obvious before the histogram equalization. This is mainly because the gray distribution range of the foreground is too wide. This can be seen from the histogram. . T o With T b They are 89 and 108, and Tf is 29. Figure 5 (a) and 5(b) respectively show the Otsu threshold T o And the threshold T obtained by the algorithm of the present invention f The segmentation result. The comparison shows that the present invention correctly separates the object and the background, but the Otsu algorithm fails, which again verifies the effect of the present algorithm on images with significant differences in the front background variance.

Embodiment example 3

[0051] Such as Image 6 (a) and 7(a) are the material images and text images segmented by Otsu's threshold method. The extracted objects are too small and too large, and Image 6 (b) and 7(b) are the segmentation results of the present invention, and the images are correctly separated. The segmentation of the two images also shows two application prospects of the present invention, which are binary image processing with uneven illumination and material image processing.

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 robust global threshold segmentation method. The method includes the steps that step1, a threshold value To of a gray-level image is calculated through an Otsu threshold value method; step2, a balance degree factor eta is defined through the variance of a foreground and a background of the gray-level image, and a threshold value Tb with the balance degree factor as the criterion is acquired; step3, a deviation degree lambda and a new optimal criterion zeta are defined according to Tb and To, and a segmentation result is generated. By means of the method, the optimal threshold value can be obtained under the condition that the probability distribution difference between the foreground and the background of the grey-level image is obvious or under general conditions.

Description

Technical field [0001] The present invention relates to the technical field of image processing, in particular to a robust global threshold segmentation method, which can avoid the failure of the most widely used Otsu threshold method when the background image gray-scale variance difference is very large before segmentation, and can also Under normal circumstances, the same results as the Otsu method are obtained, and it is a new threshold scheme that is better than the Otsu threshold method. Background technique [0002] Current image segmentation methods can be divided into pixel or region-based methods. For example, the threshold method is to binarize the image by obtaining a gray-scale threshold, and the region growth law uses initial seed points and gray-scale information-based growth criteria to obtain segmentation contours, etc. ; Boundary-based method, calculating strong edges through the edges of various gradient operators such as sobel, and then connecting the edges int...

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): G06T7/136
Inventor 龙鹏鲁华祥边昳徐露露王俭陈旭龚国良金敏陈刚
Owner INST OF SEMICONDUCTORS - CHINESE ACAD OF SCI
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