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

A Fast Segmentation Method for Active Contour Model Images Based on Gray Scale Morphological Energy Method

An active contour model and gray-scale morphology technology, applied in the field of image processing, can solve the problems of sensitive initial contour selection, low segmentation efficiency, time-consuming calculation process, etc., achieve insensitive shape and position settings, and improve curve evolution efficiency. Effect

Active Publication Date: 2019-03-29
合肥神牧信息科技有限公司
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In order to solve the problem of image segmentation falling into a local minimum, Chan et al. proposed to convert a non-convex energy function into a convex energy function and then solve it ((T.Chan, S.Esedoglu, M.Nikolova, Algorithms for finding global minimizers of image segmentation and denoisingmodels, 2006)), but the calculation process is time-consuming
To sum up, the existing active contour models based on local fitting energy are sensitive to initial contour selection and have low segmentation efficiency.

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 Fast Segmentation Method for Active Contour Model Images Based on Gray Scale Morphological Energy Method
  • A Fast Segmentation Method for Active Contour Model Images Based on Gray Scale Morphological Energy Method
  • A Fast Segmentation Method for Active Contour Model Images Based on Gray Scale Morphological Energy Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0081] In this experiment, the actual figure 2 As the grayscale image I to be segmented, its size is 168*158 pixels. The parameters in the experiment are set as follows: a 0 =1, ε=1, μ=2, υ=0.2×255 2 ,λ 1 =λ 2 =1, Δt=0.1, n=300, the structural element b is a disk with a radius of 5. The initial evolution curve is as image 3 As shown, its shape is a square with a centroid of (75, 95) and a side length of 40. Execute step 1) to step 8), Figure 4 is the result graph when iterating 10 times, Figure 5 is the result graph when iterating 50 times, Image 6 is the final segmentation result map when iterating 300 times, from Image 6 It can be seen that the present invention can effectively segment images with uneven gray levels, and has high accuracy.

Embodiment 2

[0083] Figure 7 ~ Figure 10 is processed using the classic RSF model figure 2 The obtained final segmentation result map, the parameters of the RSF model are set as follows: a 0 =1, ε=1, μ=2, υ=0.05×255 2 ,λ 1 =λ 2 = 1, Δt = 0.1, scale parameter σ = 3, the white square curve represents the initial profile, Figure 7 ~ Figure 10 The initial contours of are all squares with a side length of 40, but the positions of the centroids are different, respectively (85, 90), (35, 85), (75, 40) and (75, 130), and the white irregular curve indicates the final segmentation result.

[0084] Figure 11 to Figure 14 is to use the present invention to process figure 2 The final segmentation result figure that obtains, each parameter of the present invention is set as follows: a 0 =1, ε=1, μ=2, υ=0.2×255 2 ,λ 1 =λ 2 =1, Δt=0.1, the structural element b is a disk with a radius of 5. The white square curve represents the initial contour, Figure 11 to Figure 14 initial profile with...

Embodiment 3

[0088] Figure 15 ~ Figure 22 It uses the present invention to continuously process 30 frames of image sequences and display the final segmentation results of the first frame, the fifth frame, the ninth frame, the 13th frame, the 17th frame, the 21st frame, the 24th frame, and the 29th frame , the size of each frame of pictures is 192*240 pixels, and each parameter of the present invention is set as follows: a 0 =1, ε=1, μ=2, υ=0.4×255 2 ,λ 1 =λ 2 = 1, Δt = 0.1, the structural element b is a disk with a radius of 5, which is displayed once per frame, and the time is 10.034 seconds. The initial outline of the first frame is a square with a side length of 40, and the centroid is at (75, 130), the white square curve represents the initial outline of the first frame of pictures, and then each frame of pictures is the final segmentation curve of the previous frame as the initial outline of this frame, the number of iterations n=150 of the first frame of pictures, and then each ...

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 an ACM (Active Contour Model) image rapid segmentation method based on a gray scale morphological energy method. The segmentation method comprises the steps: firstly carrying out the gray scale corrosion and expansion calculation of an original image through a structural element, and obtaining a corrosion image and an expansion image; secondly building regional fitting energy comprising an evolution curve, the corrosion image and the expansion image, and introducing a level set function to the regional fitting energy to represent the evolution curve; adding distance rule item energy and length constraint item energy, and forming a total energy functional related with the level set function; finally solving the minimum value of the total energy functional through a steepest descent method, obtaining an evolution equation of the level set, continuously iterating the evolution equation through a finite difference method till the level set function reaches a stable state, and selecting points on a zero level set to form a segmented profile. The method can quickly and precisely segment an image with non-uniform gray scale, and is insensitive to an initial evolution curve. Moreover, the segmentation efficiency of the method is better than a conventional ACM based on local region fitting.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a method for rapidly segmenting an active contour model image based on a gray-scale morphology energy method. Background technique [0002] The image segmentation method based on the active contour model has been widely used in the field of image processing. Active contour models are mainly divided into two categories: edge-based and region-based active contour models. The edge-based active contour model mainly relies on the boundary indicator function, and makes the evolution curve stop on the target boundary according to the change of the gradient information of the target boundary, but for weak or discontinuous edges, the segmentation effect is not ideal. The region-based active contour model uses image region descriptors to achieve image partitioning, the most representative of which is the Piecewise Constant (PC) model proposed by Chan and Vese, which i...

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): G06T7/10
CPCG06T2207/10004G06T2207/10016G06T2207/20116
Inventor 饶秀勤肖林芳应义斌
Owner 合肥神牧信息科技有限公司
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