CN102496156A - Medical image segmentation method based on quantum-behaved particle swarm cooperative optimization - Google Patents
Medical image segmentation method based on quantum-behaved particle swarm cooperative optimization Download PDFInfo
- Publication number
- CN102496156A CN102496156A CN2011103665874A CN201110366587A CN102496156A CN 102496156 A CN102496156 A CN 102496156A CN 2011103665874 A CN2011103665874 A CN 2011103665874A CN 201110366587 A CN201110366587 A CN 201110366587A CN 102496156 A CN102496156 A CN 102496156A
- Authority
- CN
- China
- Prior art keywords
- individual
- optimum
- dimension
- value
- fitness function
- 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.)
- Granted
Links
Images
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
A medical image segmentation method based on quantum-behaved particle swarm cooperative optimization mainly solves the problem that increasing of categories in the prior art of image segmentation results in overlong segmentation time, local optimum of segmentation results and low segmentation precision within bearable time. The technical scheme includes step one, reading in medical images to obtain a matrix; step two, initializing population; step three, obtaining individual optimum and global optimum; step four, generating new individuals; step five, generating new individual optimum and global optimum; step six, judging whether the current iteration meets the maximum iteration or not, if yes, performing the step seven, if not, returning to the step four; step seven, performing image segmentation; and step eight, outputting segmented image matrix. The Monte Carlo method is used for multiple measurement during image segmentation threshold valuating, cooperation strategy is used for individuals obtained by multiple measurement, and accordingly the medical image segmentation method has the advantage of quickness in obtaining of ideal segmentation results and can be used for multi-threshold segmentation of medical images.
Description
Technical field
The invention belongs to technical field of image processing, further relate to a kind of of medical image segmentation technical field based on collaborative quanta particle group optimizing method.The present invention can be used for medical image such as CT image, MRI image, B ultrasonic image etc. are cut apart, compares so that realize pathology image and the normal picture of organ, and then the lesion degree of analysis organ.
Background technology
The technology that the image that medical image processing and analysis are to use computing machine that medical imaging device is collected is handled and analyzed, it can assist the doctor to diagnose more accurately.The research contents that medical image processing and analytical technology relate to has: medical image segmentation, medical figure registration, three-dimensional visualization, computer-aided diagnosis and tele-medicine etc.Wherein medical image segmentation is the basic premise of other treatment technologies, is bringing into play increasing effect at aspects such as medical research, clinical diagnosis, pathological analysis and treatments.Medical image segmentation is meant the zone or the border of from medical image, extracting tissue of interest, the tissue that is extracted can be distinguished with its hetero-organization significantly come.
Image segmentation is exactly technology and the process that is divided into image several zones specific, that have peculiar property and proposes interesting target.Below existing image partition method mainly divides several types: based on the dividing method of threshold value, based on the dividing method in zone, based on the dividing method at edge and based on dividing method of particular theory etc.Threshold segmentation method can be divided into single threshold dividing method and many threshold segmentation methods according to the characteristics of image itself.
People such as Li Dawei, Jiang Pengyuan has proposed the single threshold dividing method (" mapping science ", 2010 1 phases, 35 (1)) based on the OTSU evaluation function in " a kind of based on the image partition method that improves the OTSU evaluation function ".The OTSU method of being mentioned in the literary composition is with a type spacing variance evaluation function evaluation analysis result, is applicable to image is carried out binaryzation, promptly is divided into two types; Characteristics are simple and easy to use exactly; But its shortcoming that still exists is that along with cutting apart increasing of classification, the time increases too fast; Can bear the shortcoming that can not reach correct segmentation result in the time range, influence the effect of many Threshold Segmentation thus.
People such as Sun Jun have proposed the OTSU image partition method based on quanta particle swarm optimization in " Multilevel thresholding for image segmentation through an improvedquantum-behaved particle swarm algorithm " (IEEE Transactions on Instrumentation andMeasurement 59 (2010) 934-946).This method combines a kind of quanta particle swarm optimization and carries out image segmentation with traditional OTSU method; Optimizing process no longer is exhaustive search; Thereby time complexity has reduced on many Threshold Segmentation; But similar with other intelligent algorithms, this moment based on the image partition method of quanta particle swarm optimization in cutting procedure, stagnant evolution when still existing the segmentation result that obtains to acquire a certain degree and be absorbed in the shortcoming of a local optimum state.
Summary of the invention
The objective of the invention is to overcome the deficiency of above-mentioned prior art, proposed a kind of collaborative quantum particle swarm method that is used for the medical image Threshold Segmentation, in obtaining the individual evolutionary process of optimal particle, its update strategy is improved, performance is improved.
The thinking that the present invention realizes is to adopt the quantum particle swarm update strategy in the image segmentation process, and utilize the OTSU interpretational criteria to estimate individual quality; In the individual renewal process; Make full use of the uncertainty of quanta fluctuation in the quantum mechanics, particle utilizes DSMC to take multiple measurements when upgrading, and in order effectively to utilize the information of individual each dimension; Further the individuality that repeatedly measures is taked coordination strategy; Finally obtain the individuality expected, this moment, each individual dimension information was exactly the threshold value of split image, carried out image segmentation according to this threshold value at last.
The concrete steps that the present invention realizes comprise as follows:
(1) reads in medical image, obtain matrix, from matrix, obtain minimum gradation value and maximum gradation value;
(2) initialization population: between minimum gradation value to maximum gradation value, produce an integer at random as the first individual dimension of population; Tie up to producing an integer between the maximum gradation value at random first as the second individual dimension of population; And the like; The last dimension that initialization obtains is accomplished the individual initialization of all populations to producing each dimension between the maximum gradation value at random;
(3) obtain individual optimum and global optimum
3a) utilize the OSTU method to obtain the class spacing variance of medical image;
3b) class spacing variance is got opposite number and obtain fitness function;
3c) the population individuality is updated to fitness function and obtains the individual fitness function value of population;
3d) selecting the minimum individuality of fitness value in the population individuality obtains individual optimum;
3e) select the minimum individuality of fitness value in the individual optimum and obtain global optimum;
(4) produce new individuality
4a) each individual optimum repeatedly observation through wave function principle and Monte Carlo obtains five individuals; According to quantum particle swarm more new formula obtain first the dimension; Again according to quantum particle swarm more the numerical value that obtains of new formula greater than under the situation of first dimension with this numerical value as second dimension; Obtain second dimension otherwise first dimension is added 1, produce each dimension by that analogy;
4b) five individuals are updated to the fitness function value that fitness function obtains five individuals;
4c) from five individuals, select the minimum individuality of fitness function value as new individual;
(5) produce new individual optimum and global optimum
5a) each dimension data with each dimension data in the new individuality and all the other four individuals correspondence positions exchanges, and obtains an interim individuality;
5b) interim individuality is updated to fitness function and obtains interim individual fitness function value;
5c) if interim individual fitness function value is littler than newly individual fitness function value, then new individual with interim individual replacement, otherwise, new individual constant;
5d) if newly individual fitness function value is littler than individual optimum fitness function value, then use the new individual individual optimum that replaces, otherwise, individual optimum constant;
5e) if individual optimum fitness function value is littler than the fitness function value of global optimum, then replace global optimum, otherwise global optimum is constant with individuality is optimum;
(6) judge whether the current iteration number of times satisfies maximum iteration time,, obtain final global optimum as satisfying, otherwise, step (4) returned;
(7) carry out image segmentation: each dimension data with global optimum is cut apart the image array that obtains as threshold value, the image array after obtaining cutting apart;
(8) output of the image array after will cutting apart.
The present invention compared with prior art has following advantage:
The first, the present invention adopts the quantum particle swarm update strategy when carrying out image segmentation; And utilize the OTSU interpretational criteria to estimate individual quality; When having overcome prior art medical image segmentation branch multiclass, along with increasing of classification, the time increases too fast; Can bear the shortcoming that can not reach correct segmentation result in the time range, segmentation precision improves greatly.
Second; When the present invention optimizes class spacing variance in image segmentation; Made full use of the uncertainty of quanta fluctuation in the quantum mechanics in the evolutionary process, utilize DSMC to carry out repeatedly measuring, stagnant evolution when having overcome segmentation result that prior art obtains and acquiring a certain degree and be absorbed in the shortcoming of the state of a local optimum; Can effectively jump out local optimum, obtain better segmentation result.
The 3rd; The present invention asks in the process of threshold value in image segmentation, and the individuality that repeatedly measures has been taked coordination strategy, has made full use of each dimension information of particle; Stagnant evolution when more a step has overcome segmentation result that prior art obtains and acquires a certain degree and be absorbed in the shortcoming of the state of a local optimum; The chance of jumping out local optimum increases, and it is shorter to reach the desired result required time, thereby makes that segmentation precision further improves under the same terms.
Description of drawings
Fig. 1 is a process flow diagram of the present invention;
Fig. 2 is simulated effect figure of the present invention.
Embodiment
Below in conjunction with accompanying drawing the present invention is done further description.
With reference to Fig. 1, concrete performing step of the present invention is following:
Step 1 is read in medical image, obtains matrix, from matrix, obtains minimum gradation value and maximum gradation value;
Step 2; Initialization population: between minimum gradation value to maximum gradation value, produce an integer at random as the first individual dimension of population; Tie up to producing an integer between the maximum gradation value at random first as the second individual dimension of population, and the like, the last dimension that initialization obtains is to producing each dimension between the maximum gradation value at random; Accomplish the individual initialization of all populations, individual dimension scope is 1～4.In the embodiments of the invention dimension is taken as 2.
Step 3 obtains individual optimum and global optimum
At first, utilize the OSTU method to obtain the class spacing variance of medical image, a type spacing variance is realized by following formula:
σ
2＝ω
１×ω
2×(μ
1-μ
2)
2+ω
1×ω
2×(μ
1-μ
3)
2+ω
2×ω
3×(μ
2-μ
3)
2
Wherein, σ
2Representation class spacing variance, ω
iThe probability of representing i class gray-scale value, μ
iThe mean value of representing i class gray-scale value, i=1,2,3.
Secondly, class spacing variance is got opposite number obtain fitness function;
Once more, the population individuality is updated to fitness function and obtains the individual fitness function value of population;
Then, selecting the minimum individuality of fitness value in the population individuality obtains individual optimum;
At last, select the minimum individuality of fitness value in the individual optimum and obtain global optimum.
Step 4 produces new individuality
At first; Each individual optimum repeatedly observation through wave function principle and Monte Carlo obtains five individuals; According to quantum particle swarm more new formula obtain first the dimension; Again according to quantum particle swarm more the numerical value that obtains of new formula greater than under the situation of first dimension with this numerical value as second dimension, obtain second dimension otherwise first dimension is added 1, produce each dimension by that analogy;
Quantum particle swarm more new formula is following:
Wherein, mbest is average desired positions, and M is the individual number of population, P
IdBe the optimum d dimension of t generation i individuals, i=1,2 ... M, t are current algebraically, t=1, and 2..., G max, d=1,2, G max is a maximum iteration time;
Wherein, p
i(t+1) be the i individuals optimum in t+1 generation and the random site between the global optimum, t is current algebraically, t=1, and 2..., G max, G max are maximum iteration time,
Be the random number between 0～1, P
iBe that t generation i individuals is optimum, P
g(t) be the t global optimum in generation;
X
id(t+1)＝p
id(t+1)±α×|mbest-X
id(t)|×ln(1/μ)
Wherein, X
Id(t+1) be t+1 generation i new individual d dimension, t is current algebraically, t=1, and 2..., G max, G max are maximum iteration time, p
Id(t+1) be the d dimension of random site between i individuals optimum and the global optimum in t+1 generation, X
Id(t) be t generation i new individual d dimension, mbest is average desired positions, and α and μ are the random number between 0～1.
Step 5 is according to the more little good more principle of fitness value, from x_measure
jIn pick out optimum individual as X (t+1), and each dimension of itself and other four individuals is cooperated i=1; 2 ... n, the concrete steps of cooperation for each dimension of X (t+1) with each corresponding dimension of all the other four individuals exchange; Obtain X ' (t+1), estimate X ' (t+1) with fitness this moment, if than the good X (t+1) that then replaces of X (t+1); Be that this one dimension information replaces this original one dimension information, otherwise do not replace, finally obtain pbest
i(t+1) and gbest (t+1).
Step 6 judges whether to satisfy stop condition, obtains final gbest as satisfying, otherwise returns (4), and the stop condition here is an algebraically, and the scope of maximum iteration time is 100～1000, in the embodiment of the invention maximum iteration time is taken as 500.
Step 7 is a threshold value with the value of each dimension of gbest, the image that reads in is cut apart the image after finally obtaining cutting apart.
At first, global optimum is individual two-dimensional data is as two threshold values;
Then, be critical point with the data in the image array with three data image array after obtaining cutting apart that is divided three classes with two threshold values that obtain.
Effect of the present invention can be described further through following emulation experiment.
Accompanying drawing 2 (a) is the medical science stomach CT image that the present invention uses in emulation experiment; Accompanying drawing 2 (b) is to use the image of the present invention Fig. 2 (a) in emulation experiment to cut apart the segmentation result figure that obtains in the emulation experiment, and accompanying drawing 2 (c) is to use the collaborative quantum particle swarm method of prior art the image of Fig. 2 (a) to be cut apart the segmentation result figure that obtains in the emulation experiment.
1, simulated conditions
Emulation of the present invention is under the software environment of the hardware environment of the Pentium of dominant frequency 2.5GHZ Dual_Core CPU E5200, internal memory 2GB and MATLAB R2009a, to carry out.Test used image and derive from image library commonly used, be stomach CT figure.Experiment parameter is set to: the classification number is taken as 3, and population number is taken as 20, measures number of times and is taken as 5, and iterations is 500.
2, experiment content
Medical image and the segmentation result figure of Fig. 2 for using in the emulation experiment of the present invention, this medical image is the CT image from stomach, comprises heart, sustainer, lung, the wall of the chest and backbone, the image size is 512 * 512.
With the present invention and prior art image partition method, respectively stomach CT image 200.1,201.10 and 201.86 is cut apart based on collaborative quanta particle swarm optimization.
According to the present invention with the cutting procedure of prior art based on the image partition method of collaborative quanta particle swarm optimization; After cutting apart completion; The class spacing variance that can obtain final segmentation result is estimated segmentation result figure, and evaluation result is as shown in the table, wherein; CQPSO represents method of the present invention, and the sunCQPSO representative is based on the image partition method of collaborative quanta particle swarm optimization.
Emulation experiment of the present invention has also carried out being shown as Fig. 2 with the segmentation result figure of first width of cloth image in the following table.
3. interpretation
Cutting apart the data that obtain through the invention described above method and art methods can see; The class spacing variance that the inventive method obtains is all cut apart the class spacing variance that obtains greater than art methods; According to the principle that OTSU interpretational criteria class spacing variance is the bigger the better, segmentation precision has improved under the same terms of the present invention, can obtain better segmentation result; Show simultaneously obtain identical result cut apart figure the time; The present invention needs iterations still less, and it is shorter then to reach the desired result required time, so segmentation performance of the present invention is better.
Fig. 2 is the segmentation result figure to medical image; Simulation result by Fig. 2 (b) and Fig. 2 (c) can be seen; Segmentation result Fig. 2 (b) that the present invention obtains can effectively be cut apart organ clear; And edge's segmentation result is better, compares trulyr with former figure, shows that the inventive method can obtain segmentation result more accurately in medical image.Organ is cut apart fuzzylyyer among segmentation result Fig. 2 (c) that prior art obtains based on the image partition method of collaborative quanta particle swarm optimization, and in the situation of the many local wrong branches of organ, is difficult to distinguish different organs.
Claims (6)
1. the medical image cutting method based on collaborative quanta particle swarm optimization comprises the steps:
(1) reads in medical image, obtain matrix, from matrix, obtain minimum gradation value and maximum gradation value;
(2) initialization population: between minimum gradation value to maximum gradation value, produce an integer at random as the first individual dimension of population; Tie up to producing an integer between the maximum gradation value at random first as the second individual dimension of population; And the like; The last dimension that initialization obtains is accomplished the individual initialization of all populations to producing each dimension between the maximum gradation value at random;
(3) obtain individual optimum and global optimum
3a) utilize the OSTU method to obtain the class spacing variance of medical image;
3b) class spacing variance is got opposite number and obtain fitness function;
3c) the population individuality is updated to fitness function and obtains the individual fitness function value of population;
3d) selecting the minimum individuality of fitness value in the population individuality obtains individual optimum;
3e) select the minimum individuality of fitness value in the individual optimum and obtain global optimum;
(4) produce new individuality
4a) each individual optimum repeatedly observation through wave function principle and Monte Carlo obtains five individuals; According to quantum particle swarm more new formula obtain first the dimension; Again according to quantum particle swarm more the numerical value that obtains of new formula greater than under the situation of first dimension with this numerical value as second dimension; Obtain second dimension otherwise first dimension is added 1, produce each dimension by that analogy;
4b) five individuals are updated to the fitness function value that fitness function obtains five individuals;
4c) from five individuals, select the minimum individuality of fitness function value as new individual;
(5) produce new individual optimum and global optimum
5a) each dimension data with each dimension data in the new individuality and all the other four individuals correspondence positions exchanges, and obtains an interim individuality;
5b) interim individuality is updated to fitness function and obtains interim individual fitness function value;
5c) if interim individual fitness function value is littler than newly individual fitness function value, then new individual with interim individual replacement, otherwise, new individual constant;
5d) if newly individual fitness function value is littler than individual optimum fitness function value, then use the new individual individual optimum that replaces, otherwise, individual optimum constant;
5e) if individual optimum fitness function value is littler than the fitness function value of global optimum, then replace global optimum, otherwise global optimum is constant with individuality is optimum;
(6) judge whether the current iteration number of times satisfies maximum iteration time,, obtain final global optimum as satisfying, otherwise, step (4) returned;
(7) carry out image segmentation: each dimension data with global optimum is cut apart the image array that obtains as threshold value, the image array after obtaining cutting apart;
(8) output of the image array after will cutting apart.
2. the medical image cutting method based on collaborative quanta particle swarm optimization according to claim 1 is characterized in that the described individual dimension of step (2) is 1～4.
3. the medical image cutting method based on collaborative quanta particle swarm optimization according to claim 1 is characterized in that step 3a) described OSTU method realizes by following formula:
σ
2＝ω
1×ω
2×(μ
1-μ
2)
2+ω
1×ω
3×(μ
1-μ
3)
2+ω
2×ω
3×(μ
2-μ
3)
2
Wherein, σ
2Representation class spacing variance, ω
iThe probability of representing i class gray-scale value, μ
iThe mean value of representing i class gray-scale value, i=1,2,3.
4. the medical image cutting method based on collaborative quanta particle swarm optimization according to claim 1 is characterized in that step 4a) it is described that more new formula is following according to quantum particle swarm:
The first step, calculate average desired positions by following formula:
Wherein, mbest is average desired positions, and M is the individual number of population, P
IdBe the optimum d dimension of t generation i individuals, i=1,2 ... M, t are current algebraically, t=1, and 2..., G max, d=1,2, G max is a maximum iteration time;
In second step, try to achieve the study variable by following formula:
L(t+1)＝2×α×|mbest-X
i(t)|
Wherein, L (t+1) is the study variable, and t is current algebraically, t=1, and 2..., G max, G max are maximum iteration time, and α is the random number between 0～1, and mbest is average desired positions, X
i(t) be the i individuals;
The 3rd goes on foot, and tries to achieve the dimension of new individuality by following formula:
X
id(t+1)＝P
id(t)±α×|mbest-X
id(t)|×ln(1/μ)
Wherein, X
i(t+1) be t+1 generation i new individual d dimension, t is current algebraically, t=1, and 2..., G max, G max are maximum iteration time, P
IdBe the optimum d dimension of t generation i individuals, α and μ are the random number between 0～1.
5. the medical image cutting method based on collaborative quanta particle swarm optimization according to claim 1 is characterized in that the scope of the said maximum iteration time of step (6) is 100～1000.
6. the medical image cutting method based on collaborative quanta particle swarm optimization according to claim 1 is characterized in that the step of the said image segmentation of step (7) is:
The first step, the two-dimensional data that global optimum is individual is as two threshold values;
Second step was critical point with the data in the image array with three data image array after obtaining cutting apart that is divided three classes with two threshold values that obtain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011103665874A CN102496156B (en) | 2011-11-17 | 2011-11-17 | Medical image segmentation method based on quantum-behaved particle swarm cooperative optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011103665874A CN102496156B (en) | 2011-11-17 | 2011-11-17 | Medical image segmentation method based on quantum-behaved particle swarm cooperative optimization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102496156A true CN102496156A (en) | 2012-06-13 |
| CN102496156B CN102496156B (en) | 2013-11-20 |
Family
ID=46187978
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011103665874A Expired - Fee Related CN102496156B (en) | 2011-11-17 | 2011-11-17 | Medical image segmentation method based on quantum-behaved particle swarm cooperative optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102496156B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102903112A (en) * | 2012-09-29 | 2013-01-30 | 江南大学 | Small target detection-orientated image threshold segmentation method adopting fast kernel density estimation |
| CN102903113A (en) * | 2012-10-08 | 2013-01-30 | 南京邮电大学 | Multi-threshold image segmentation method based on cooperative quantum particle swarm algorithm |
| CN103136759A (en) * | 2013-03-26 | 2013-06-05 | 长春工业大学 | White matter magnetic resonance imaging (MRI) picture segmentation method based on bionic optimization |
| CN103544695A (en) * | 2013-09-28 | 2014-01-29 | 大连理工大学 | Efficient medical image segmentation method based on game framework |
| CN103985112A (en) * | 2014-03-05 | 2014-08-13 | 西安电子科技大学 | Image segmentation method based on improved multi-objective particle swarm optimization and clustering |
| CN105405136A (en) * | 2015-11-04 | 2016-03-16 | 南方医科大学 | Self-adaptive spinal CT image segmentation method based on particle swarm optimization |
| CN105631456A (en) * | 2015-12-15 | 2016-06-01 | 安徽工业大学 | Particle swarm optimization ITTI model-based white cell region extraction method |
| CN106023205A (en) * | 2016-05-23 | 2016-10-12 | 杭州健培科技有限公司 | Medical image segmentation method based on simplified PSO (Particle Swarm Optimization) and 2D maximum entropy threshold |
| CN106296704A (en) * | 2016-08-16 | 2017-01-04 | 中国科学技术大学 | Universal image partition method |
| CN107067419A (en) * | 2017-04-16 | 2017-08-18 | 江西理工大学 | The method for registering images of application enhancements gravitation search |
| CN107103609A (en) * | 2017-04-17 | 2017-08-29 | 上海电力学院 | Niblack power equipment Infrared Image Segmentations based on particle group optimizing |
| CN107590900A (en) * | 2017-09-15 | 2018-01-16 | 深圳怡化电脑股份有限公司 | A kind of crown word number image acquiring method, device, equipment and storage medium |
| CN110553376A (en) * | 2019-09-18 | 2019-12-10 | 福州大学 | CQPSO algorithm-based VAV air-conditioning system temperature control method |
| CN111899250A (en) * | 2020-08-06 | 2020-11-06 | 罗春华 | Remote disease intelligent diagnosis system based on block chain and medical image |
| CN113269756A (en) * | 2021-05-28 | 2021-08-17 | 长春大学 | Retina blood vessel segmentation method and device based on multi-scale matched filtering and particle swarm optimization |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101699514A (en) * | 2009-10-30 | 2010-04-28 | 西安电子科技大学 | Immune clone quantum clustering-based SAR image segmenting method |
| CN101887584A (en) * | 2010-07-07 | 2010-11-17 | 清华大学 | Fitness random search behavior-based multi-threshold image segmentation method |
-
2011
- 2011-11-17 CN CN2011103665874A patent/CN102496156B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101699514A (en) * | 2009-10-30 | 2010-04-28 | 西安电子科技大学 | Immune clone quantum clustering-based SAR image segmenting method |
| CN101887584A (en) * | 2010-07-07 | 2010-11-17 | 清华大学 | Fitness random search behavior-based multi-threshold image segmentation method |
Non-Patent Citations (3)
| Title |
|---|
| HAO GAO ET AL: "Multilevel Thresholding for Image Segmentation Through an Improved Quantum-Behaved Particle Swarm Algorithm", 《IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT》, vol. 59, no. 4, 30 April 2010 (2010-04-30) * |
| 冯斌等: "基于量子粒子群算法的Ostu图像阈值分割", 《计算机工程与设计》, vol. 29, no. 3, 16 July 2008 (2008-07-16) * |
| 高浩: "粒子群算法及其在图像分割中的应用与研究", 《中国博士学位论文全文数据库 信息科技辑》, no. 12, 31 December 2010 (2010-12-31) * |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102903112B (en) * | 2012-09-29 | 2015-01-07 | 江南大学 | Small target detection-orientated image threshold segmentation method adopting fast kernel density estimation |
| CN102903112A (en) * | 2012-09-29 | 2013-01-30 | 江南大学 | Small target detection-orientated image threshold segmentation method adopting fast kernel density estimation |
| CN102903113A (en) * | 2012-10-08 | 2013-01-30 | 南京邮电大学 | Multi-threshold image segmentation method based on cooperative quantum particle swarm algorithm |
| CN103136759A (en) * | 2013-03-26 | 2013-06-05 | 长春工业大学 | White matter magnetic resonance imaging (MRI) picture segmentation method based on bionic optimization |
| CN103544695A (en) * | 2013-09-28 | 2014-01-29 | 大连理工大学 | Efficient medical image segmentation method based on game framework |
| CN103544695B (en) * | 2013-09-28 | 2015-12-23 | 大连理工大学 | A kind of efficiently based on the medical image cutting method of game framework |
| CN103985112B (en) * | 2014-03-05 | 2017-05-10 | 西安电子科技大学 | Image segmentation method based on improved multi-objective particle swarm optimization and clustering |
| CN103985112A (en) * | 2014-03-05 | 2014-08-13 | 西安电子科技大学 | Image segmentation method based on improved multi-objective particle swarm optimization and clustering |
| CN105405136A (en) * | 2015-11-04 | 2016-03-16 | 南方医科大学 | Self-adaptive spinal CT image segmentation method based on particle swarm optimization |
| CN105631456B (en) * | 2015-12-15 | 2018-11-30 | 安徽工业大学 | A kind of leucocyte method for extracting region based on particle group optimizing ITTI model |
| CN105631456A (en) * | 2015-12-15 | 2016-06-01 | 安徽工业大学 | Particle swarm optimization ITTI model-based white cell region extraction method |
| CN106023205A (en) * | 2016-05-23 | 2016-10-12 | 杭州健培科技有限公司 | Medical image segmentation method based on simplified PSO (Particle Swarm Optimization) and 2D maximum entropy threshold |
| CN106296704A (en) * | 2016-08-16 | 2017-01-04 | 中国科学技术大学 | Universal image partition method |
| CN106296704B (en) * | 2016-08-16 | 2019-02-01 | 中国科学技术大学 | Universal image partition method |
| CN107067419A (en) * | 2017-04-16 | 2017-08-18 | 江西理工大学 | The method for registering images of application enhancements gravitation search |
| CN107103609A (en) * | 2017-04-17 | 2017-08-29 | 上海电力学院 | Niblack power equipment Infrared Image Segmentations based on particle group optimizing |
| CN107103609B (en) * | 2017-04-17 | 2019-06-11 | 上海电力学院 | Niblack power equipment Infrared Image Segmentation based on particle group optimizing |
| CN107590900A (en) * | 2017-09-15 | 2018-01-16 | 深圳怡化电脑股份有限公司 | A kind of crown word number image acquiring method, device, equipment and storage medium |
| CN110553376A (en) * | 2019-09-18 | 2019-12-10 | 福州大学 | CQPSO algorithm-based VAV air-conditioning system temperature control method |
| CN111899250A (en) * | 2020-08-06 | 2020-11-06 | 罗春华 | Remote disease intelligent diagnosis system based on block chain and medical image |
| CN113269756A (en) * | 2021-05-28 | 2021-08-17 | 长春大学 | Retina blood vessel segmentation method and device based on multi-scale matched filtering and particle swarm optimization |
| CN113269756B (en) * | 2021-05-28 | 2024-02-27 | 长春大学 | Method and device for optimizing retinal vessel segmentation based on multi-scale matched filtering and particle swarm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102496156B (en) | 2013-11-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102496156B (en) | Medical image segmentation method based on quantum-behaved particle swarm cooperative optimization | |
| CN105009174B (en) | Method and system for assessing tissue fibrosis | |
| CN108257135A (en) | The assistant diagnosis system of medical image features is understood based on deep learning method | |
| EP2027566B1 (en) | Automatic recognition of preneoplastic anomalies in anatomic structures based on an improved region-growing segmentation, and computer program therefor | |
| CN111179237B (en) | Liver and liver tumor image segmentation method and device | |
| CN109360208A (en) | A kind of medical image cutting method based on one way multitask convolutional neural networks | |
| CN112568923B (en) | X-ray phase contrast image extraction method, device, terminal and storage medium | |
| CN106023205A (en) | Medical image segmentation method based on simplified PSO (Particle Swarm Optimization) and 2D maximum entropy threshold | |
| CN110874860A (en) | Target extraction method of symmetric supervision model based on mixed loss function | |
| CN109215040A (en) | A kind of tumor of breast dividing method based on multiple dimensioned weighting study | |
| Fashandi et al. | An investigation of the effect of fat suppression and dimensionality on the accuracy of breast MRI segmentation using U‐nets | |
| Kamiya | Deep learning technique for musculoskeletal analysis | |
| CN110459303B (en) | Medical image abnormity detection device based on depth migration | |
| CN109034259A (en) | Image control system, method, terminal, the storage medium of carcinoma intervention device | |
| Do et al. | Supporting thyroid cancer diagnosis based on cell classification over microscopic images | |
| CN110232695A (en) | Left ventricle image partition method and system based on hybrid mode image | |
| CN113283465B (en) | Diffusion tensor imaging data analysis method and device | |
| CN111127636B (en) | Intelligent complex intra-articular fracture desktop-level three-dimensional diagnosis system | |
| CN114511564A (en) | Image analysis method for breast cancer residual tumor load based on DCE-MRI | |
| Song et al. | An improved confidence connected liver segmentation method based on three views of CT images | |
| Chang et al. | CT manifestations of gallbladder carcinoma based on neural network | |
| CN113313714A (en) | Coronary artery OCT image lesion plaque segmentation method based on improved U-Net network | |
| CN113205472A (en) | Cross-modal MR image mutual generation method based on cyclic generation countermeasure network cycleGAN model | |
| JP2004174220A (en) | Apparatus and method for processing image and recording medium for storing program used for causing computer to execute the method | |
| Elmasry et al. | Graph partitioning based automatic segmentation approach for ct scan liver images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20131120 Termination date: 20191117 |