CN106447678A - Medical image segmentation method based on regional mixed movable contour model - Google Patents
Medical image segmentation method based on regional mixed movable contour model Download PDFInfo
- Publication number
- CN106447678A CN106447678A CN201610899963.9A CN201610899963A CN106447678A CN 106447678 A CN106447678 A CN 106447678A CN 201610899963 A CN201610899963 A CN 201610899963A CN 106447678 A CN106447678 A CN 106447678A
- Authority
- CN
- China
- Prior art keywords
- image
- medical image
- region
- method based
- segmentation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20021—Dividing image into blocks, subimages or windows
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30008—Bone
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30016—Brain
Abstract
The invention discloses a medical image segmentation method based on a regional mixed movable contour model. In the invention, the segmentation method is provided mainly aiming at a medical class of images. The method is characterized by (1) establishing the movable contour model based on a mixed region and accelerating image target area segmentation and contour curve fitting; and (2) according to a non-convex regularization item increased because of image composition and a local cluster property of the images, maintaining a geometrical shape of the region and preventing an edge from generating a over-smoothing phenomenon. The method is simple and easy to operate, a target area in a medical image can be accurately segmented, and a convergence speed is rapid and accuracy is good.
Description
Technical field：
The present invention relates to computer vision field, specifically a kind of side that target in medical image is split
Method.
Background technology：
Continuous development with computer technology and progress, people are to the application demand of computer vision aspect also increasingly
Many.Such as, the products such as the digital camera that is seen everywhere in life, digital camera, smart mobile phone and consumer's life breath
Manner of breathing closes.Target in image is carried out with detection and causes increasing concern with identification, and intelligent monitoring, man-machine interaction,
The fields such as medical treatment all have extremely wide application prospect.
Image segmentation is the important step of image procossing, and the development being successive image engineering to the research of image segmentation is played
Lay a foundation well effect.
Content of the invention：
The purpose of the present invention is for building a kind of mesh with higher using value, simple Zhen Dui medical image
Mark dividing method.
The present invention passes through the minimum of energy functional in modulus type, obtains the gradient with regard to level set function and declines equation,
And its discretization is obtained with the iterative calculation formula of level set function, rapidly realize the segmentation to image object.To different doctors
The emulation experiment learning image shows, compared with other same type algorithms, the model set up in the present invention and corresponding algorithm can
To obtain more accurate segmentation result within the shorter time.
Specific technical scheme is as follows：
(1) adopt biphase and many phase method to initialize level set function, and initialize parameters；
(2) gray average of calculated level set function each sub-regions in the picture；
(3) calculate the skew field variable of image；
(4) the new value according to iterative calculation formula calculated level set function, until obtaining the contour curve of image object,
Realize the segmentation of target.
The invention has the beneficial effects as follows:
1st, set up a kind of medical image segmentation model based on region hybrid activity profile, design corresponding partitioning algorithm；
2nd, the present invention is simple, is capable of quick medical science in the case of uneven illumination is even with imaging device imperfection
The segmentation of image object, meets requirement of real-time, is widely used, and the detection for medical image is referred to identification and applies
Value.
Target Segmentation be applied to medical image for the present invention, can provide reference for medical image engineering, to medical science figure
The effect of laying a foundation is played in the analysis of picture, identification and understanding.
Brief description：
Fig. 1 is the biphase segmentation result figure to Patella image for the present invention；
Fig. 2 is the multiphase segmentation result figure to people's brain image for the present invention；
Fig. 3 is the multiphase segmentation time to people's brain image for the present invention
Specific embodiment：
Further illustrate the flesh and blood of the present invention below in conjunction with the accompanying drawings with example, but present disclosure is not limited to
This.
Embodiment 1：
Take the brain medical image of the compressed format people that size is 193 × 255, and determine each parameter in model.Using
Four phase methods randomly initialize two level set functions, and obtain corresponding initial profile curve, thus dividing the image into many
Individual region (left figure in Fig. 2)；The skew field quantity of calculated level the set function gray average of each sub-regions and image in the picture,
Update the value of level set function, such iteration, until obtaining patellar contour curve in image, thus obtain segmentation result
(right figure in Fig. 2).
Claims (4)
1. a kind of medical image cutting method based on region hybrid activity skeleton pattern.The present invention enables mesh in medical image
Target is split, and does pretreatment and analysis work for identification target, provides reference to the process of medical image.It is characterized in that：
(1) set up the hybrid activity skeleton pattern clustering property based on curve geometry metric parameter and local gray level；
(2) energy functional in model comprises data fit term and regularization term：Data fit term by view data fit term and
Geodetic data fit term forms, it may ensure that contour curve develops towards edges of regions to be split；And regularization term is used
Carry out controlling profile curve and keep preferable shape in evolutionary process；
(3) gray average of each sub-regions and biased field in image are determined, and to the petition of surrender under the gradient of each level set function
Reach formula and carry out discretization, obtain iterative with regard to level set function, finally realize the segmentation to image object.
2. a kind of medical image cutting method based on region hybrid activity skeleton pattern according to claim 1, it is special
Levy and be：Realize the evolution to multiple curved profiles in the case of considering image topology structure, solve because of uneven illumination simultaneously
The even segmentation effect leading to the uneven image of gray scale with imaging device imperfection problem not fully up to expectations.
3. a kind of medical image cutting method based on region hybrid activity skeleton pattern according to claim 1, it is special
Levy and be：Image-region is divided into disjoint image region using some Confined outline curves, calculates each sub-regions
On image fitting function, accelerate curve evolvement speed, and ensure contour curve in an iterative process towards needing cut zone
Edge and develop.
4. a kind of medical image cutting method based on region hybrid activity skeleton pattern according to claim 1, it is special
Levy and be：Calculate length, distance and the non-convex of contour curve and they are added in image fitting function as regular terms, keep
The geometry in region, prevents produce in image segmentation process excessively to smooth phenomenon.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610899963.9A CN106447678A (en) | 2016-10-14 | 2016-10-14 | Medical image segmentation method based on regional mixed movable contour model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610899963.9A CN106447678A (en) | 2016-10-14 | 2016-10-14 | Medical image segmentation method based on regional mixed movable contour model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106447678A true CN106447678A (en) | 2017-02-22 |
Family
ID=58174881
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610899963.9A Pending CN106447678A (en) | 2016-10-14 | 2016-10-14 | Medical image segmentation method based on regional mixed movable contour model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106447678A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109472792A (en) * | 2018-10-29 | 2019-03-15 | 石家庄学院 | In conjunction with the local energy functional of local entropy and the image partition method of non-convex regular terms |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1471054A (en) * | 2002-07-26 | 2004-01-28 | 中国科学院自动化研究所 | Automatic segmentation method of multi targets based moving contour model |
| US20080030497A1 (en) * | 2005-12-08 | 2008-02-07 | Yangqiu Hu | Three dimensional modeling of objects |
| CN103679734A (en) * | 2013-12-25 | 2014-03-26 | 浙江师范大学 | Method for eyed typhoon two-dimensional surface wind field inversion on basis of SVM and PDE |
-
2016
- 2016-10-14 CN CN201610899963.9A patent/CN106447678A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1471054A (en) * | 2002-07-26 | 2004-01-28 | 中国科学院自动化研究所 | Automatic segmentation method of multi targets based moving contour model |
| US20080030497A1 (en) * | 2005-12-08 | 2008-02-07 | Yangqiu Hu | Three dimensional modeling of objects |
| CN103679734A (en) * | 2013-12-25 | 2014-03-26 | 浙江师范大学 | Method for eyed typhoon two-dimensional surface wind field inversion on basis of SVM and PDE |
Non-Patent Citations (1)
| Title |
|---|
| 林喜兰: "基于区域混合活动轮廓模型的医学图像分割", 《计算机科学》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109472792A (en) * | 2018-10-29 | 2019-03-15 | 石家庄学院 | In conjunction with the local energy functional of local entropy and the image partition method of non-convex regular terms |
| CN109472792B (en) * | 2018-10-29 | 2021-08-20 | 石家庄学院 | Local energy functional and non-convex regular term image segmentation method combining local entropy |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105139420B (en) | A kind of video target tracking method based on particle filter and perception Hash | |
| Hu et al. | 3D Pose tracking of walker users' lower limb with a structured-light camera on a moving platform | |
| CN107301657B (en) | A kind of video target tracking method considering target movable information | |
| CN109816692A (en) | A kind of motion target tracking method based on Camshift algorithm | |
| CN106651871B (en) | Automatic filling method for depth image hole | |
| CN108665485A (en) | A kind of method for tracking target merged with twin convolutional network based on correlation filtering | |
| CN102831382A (en) | Face tracking apparatus and method | |
| IN2013CH05313A (en) | ||
| CN102136142A (en) | Nonrigid medical image registration method based on self-adapting triangular meshes | |
| CN108159576B (en) | Human body chest and abdomen surface area respiratory motion prediction method in radiotherapy | |
| CN110189341A (en) | A kind of method, the method and device of image segmentation of Image Segmentation Model training | |
| CN103150723A (en) | Lymph gland detection system and method based on stomach computed tomography (CT) image with shape fitting with oval | |
| CN104835168B (en) | Quick multiphase image dividing method based on global convex optimization Variation Model | |
| CN106443662B (en) | A kind of target steady correlating method under low repetition system when velocity ambiguity | |
| CN103268603A (en) | Liver image segregation method based on hierarchy vessel tree division | |
| Wallenberg et al. | Channel coding for joint colour and depth segmentation | |
| CN106447678A (en) | Medical image segmentation method based on regional mixed movable contour model | |
| CN105069829B (en) | A kind of human body animation generation method based on more visually frequencies | |
| CN103208114A (en) | Stomach adipose tissue extraction method based on interactive segmentation | |
| CN109544565A (en) | A kind of surface of a wound image partition method based on DC-GMRF level set | |
| JP2014010717A (en) | Area division device | |
| CN108682021A (en) | Rapid hand tracking, device, terminal and storage medium | |
| CN103810480A (en) | Method for detecting gesture based on RGB-D image | |
| CN103440681A (en) | Non-contact nondestructive omnibearing three-dimensional modeling method | |
| CN109730769A (en) | A kind of skin neoplasin based on machine vision is precisely performed the operation intelligent method for tracing and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170222 |