Regional growth image segmentation method based on local information

An image segmentation and region growing technology, applied in image analysis, image enhancement, image data processing and other directions, can solve the problems of image over-segmentation or under-segmentation, uneven grayscale, and poor image segmentation effect, so as to avoid over-segmentation or under-segmentation. The effect of under-segmentation, noise insensitivity, and high segmentation accuracy

Active Publication Date: 2019-12-27
HUNAN UNIV OF SCI & TECH
View PDF4 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The existing region growing methods usually have the following disadvantages: the definition of the growth criterion is blind, it is difficult to find a suitable growth criterion in advance, and the inappropriate growth criterion will easily cause the image to be over-segmented or under-segmented; due to the existing The technology usually uses single pixel features and fixed growth criteria to segment images. Therefore, the segmentation effect is poor for images with blurred target boundaries, uneven gray levels, and rich textures; it is sensitive to noise, and it is easy to generate segmentation holes for noisy images, especially when When the seed point is specified as a noise point, this method will not be able to segment

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
  • Regional growth image segmentation method based on local information
  • Regional growth image segmentation method based on local information
  • Regional growth image segmentation method based on local information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] A region growing image segmentation method based on local information, the specific implementation steps are as follows:

[0032] (1) Manually select a seed point u in the target area of ​​the original image to figure 1 The original image shown as an example, figure 2 The black hollow point in the middle is the manually selected seed point;

[0033] (2) Calculate the average value of the image area with the seed point u as the center radius smaller than r on the three color channels of R, G, and B and standard deviation Where r is a constant greater than 0, preferably a constant between 2 and 10, and r=5 is preferred in this embodiment;

[0034] (3) Initialize i=1, k i = ω, S i ={u}, where ω is a constant greater than 0, preferably a constant between 0.2 and 2, and in this embodiment, ω=1 is preferred;

[0035] (4) Taking the seed point u as the current point, calculate its 8-neighborhood pixel set Ω u Each pixel p∈Ω in u The local color mean of with

[...

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 regional growth image segmentation method based on local information, and the method comprises the following steps: firstly selecting a seed point in an image target region;then, taking the seed point as a starting point, carrying out multi-scale region growth on the image by utilizing pixel local information and adopting growth criteria of different scales, and dynamically updating the growth criteria by utilizing R, G and B color mean values and standard deviation of an intermediate result in each region growth; and finally, calculating a probability distribution difference between two adjacent region growth results by adopting Kullback-Leibler divergence, and when a difference value is greater than a preset threshold value, taking a previous region growth result as a final segmentation result. The method is insensitive to noise, can effectively segment images with fuzzy target boundaries, uneven gray levels and rich textures, and is high in segmentation precision and robustness.

Description

technical field [0001] The invention belongs to the field of digital image processing, relates to a digital image segmentation technology, in particular to a region growing image segmentation method based on local information. [0002] technical background [0003] Region growing is an image segmentation method that aggregates pixels with similar attributes. This method first needs to specify a group of seed points (the seed point can be a single pixel or a small area), and then judge the pixels in the neighborhood of the seed point according to the growth criterion defined in advance. If the similarity between pixels satisfies the growth criterion, the neighboring pixels are marked as the target, and the neighboring pixel is used as the new seed point, and the above search and judgment process is repeated until all satisfying pixel points are included. The growing similarity criterion can be defined according to the image data type by using information such as grayscale and...

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): G06T7/11
CPCG06T7/11G06T2207/20101G06T2207/20076
Inventor 廖苗赵于前邸拴虎刘毅志刘建勋
Owner HUNAN UNIV OF SCI & TECH
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