CN107045717A - The detection method of leucocyte based on artificial bee colony algorithm - Google Patents
The detection method of leucocyte based on artificial bee colony algorithm Download PDFInfo
- Publication number
- CN107045717A CN107045717A CN201710169481.2A CN201710169481A CN107045717A CN 107045717 A CN107045717 A CN 107045717A CN 201710169481 A CN201710169481 A CN 201710169481A CN 107045717 A CN107045717 A CN 107045717A
- Authority
- CN
- China
- Prior art keywords
- nectar source
- honeybee
- leucocyte
- bee colony
- ellipse
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses the detection method of the leucocyte based on artificial bee colony algorithm, first using edge detection process image, and extract all edge pixels and be stored in vector P, then population is initialized.Calculate the fitness value in nectar source;Employ honeybee to scan for producing new nectar source, calculate its adaptive value and preferable nectar source is selected using greedy method;Honeybee sharing information is employed, follows honeybee to calculate the selected probability in nectar source using roulette principle, while according to search formula nearby fine search;Determine whether that food source continuous limit generations do not update, honeybee is investigated if then employing honeybee to be changed into, investigation honeybee search produces new nectar source;The optimal nectar source position of memory;Judge whether to reach maximum iteration, output optimal solution is terminated if reaching, finally realizes that the image of leucocyte is detected.Artificial bee colony algorithm is used for the detection of leucocyte by the present invention, while preferable detection speed is kept, improves the precision of white blood cell detection.
Description
Technical field
The invention belongs to image processing field, the side of more particularly to a kind of white blood cell detection based on artificial bee colony algorithm
Method.
Background technology
Blood of human body leucocyte is also commonly known as immunocyte, is played an important role in human immune system, in clinic
Medically can be by the quantity for observing different classes of leucocyte and modal change, so as to judge the pathology of function of human body
Situation.Therefore, the detection of leucocyte is an important indicator of blood routine detection.At present, white blood cell detection is mainly with microscope
Based on Automated hematologic analyzer, but they have respective shortcoming, such as the time length of cervical arthroplasty consuming, work
Amount is big and dull;The shortcomings of self-reacting device can not detect the change of WBC Appearance.
With the continuous progress and development in the fields such as image procossing and pattern-recognition, these technologies may apply to medical science shadow
As in, the detection in leucocyte can also be applied.White blood cell detection can be divided into two steps of rim detection and ellipses detection.
The method of ellipses detection can substantially be divided into two classes, and a class is ballot method, for example, Hough transform method and RANSAC algorithms etc.;Separately
One class is optimized algorithm, such as least square fitting, genetic algorithm, artificial bee colony algorithm;In current practical application, most
Classical method is that the image object of ellipse is detected using Hough transform, and Hough transform method lacks for oval part
Become estranged insensitive for noise, with very high accuracy of detection and robustness, but be due to that ellipse there are five own parameters, it is necessary to five
Dimension parameter space is accumulated, and causes this way because amount of calculation and memory demand are excessive without gearing to actual circumstances, through
The Hough transform of allusion quotation is simultaneously improper.And optimized algorithm can overcome these problems.Solve optimization problem method can be divided into through
Allusion quotation optimization method and intelligent optimization algorithm.Classical optimized algorithm includes gradient decline, simple row method, conjugate direction method, newton
Method etc..Classic optimisation algorithm it is complicated for some, discrete or multiple target system be difficult to have preferable optimum results.With
Classical optimization method is compared, and artificial bee colony algorithm is not almost required object function and constraint, in search procedure almost
External information is not utilized, only using fitness function as evolution foundation, the artificial intelligence being characterized with " generation+inspection " is formd
Technology.Intelligent optimization algorithm includes ant colony optimization algorithm, particle swarm optimization algorithm, ant colony algorithm, difference algorithm etc..Population is calculated
Although method fast convergence rate, is easily trapped into locally optimal solution;Although difference algorithm is difficult to be absorbed in locally optimal solution, but restrains
Speed is slow.The characteristics of artificial bee colony algorithm has small simple to operate, control parameter, high search precision and stronger robustness.
The content of the invention
There is problem in the present invention traditional white blood cell detection for more than, propose a kind of leucocyte based on artificial bee colony algorithm
Detection method, artificial bee colony algorithm can jump out locally optimal solution, the performance of globally optimal solution can be also searched well, is had
Effect improves the speed and precision of operation.
To achieve the above object, technical scheme proposed by the present invention is the detection side of the leucocyte based on artificial bee colony algorithm
Method, comprises the following steps:
Step 1:Using edge detection process image, and extract all edge pixels and be stored in vector P;
Step 2:Initialize population:Initialize nectar source xid, setting nectar source sum NP, control parameter limit, greatest iteration time
Number, t=1;
Step 3:Calculate the fitness value in nectar source;
Step 4:Employ honeybee to scan for producing new nectar source, calculate its adaptive value and using the preferable honey of greedy method selection
Source;
Step 5:Honeybee sharing information is employed, follows honeybee to calculate the selected probability in nectar source using roulette principle, while root
According to search formula nearby fine search;
Step 6:Determine whether that food source continuous limit generations do not update, if then employing honeybee to be changed into investigating honeybee, detect
Look into honeybee search and produce new nectar source;Step 8 is gone to if not;
Step 7:The optimal nectar source position of memory;
Step:8:T=t+1;Judge whether to reach maximum iteration, output optimal solution is terminated if reaching, otherwise
Repeat step 3-7;
Step 9:The image of leucocyte is detected.
Further, the rim detection in above-mentioned steps 1 specifically includes the side for extracting leukocyte cell core ellipse in picture
Edge profile, then carries out ellipses detection, extracts all edge pixels and stores into vector P, P=(P1,P2,…,Pn), n is figure
As the summation of upper all edge pixels.
The position in each nectar source represents the ellipse of a feasible solution in above-mentioned steps 2, often produce one it is new
Candidate nectar source represents a kind of new candidate's ellipse, initializes the vector in all nectar sourcesOval base
This equation can be expressed as:
ax2+bxy+cy2+ dx+ey+1=0
As five free parameters a, b, c, d, e, value meet b2Ellipse is constituted during -4ac < 0, so each is oval
It can be represented by five free parameters, in order to build ellipse, arbitrarily extract five edge pixel point p out in vector Pi1,pi2,
pi3,pi4,pi5, wherein any three points are not point-blank, an ellipse can be constituted by being combined, so each
Individual nectar source is all encoded to an ellipse Xi, when carrying out ellipses detection, arbitrary five edge pixels point difference structure in vector P
Into five systems of linear equations on parameter, it is solved, if obtained solution meets condition, using artificial bee colony algorithm pair
Solution carries out selection optimization.
The position that nectar source is initialized in above-mentioned steps 2 is:
xid=Ld+rand(0,1)(Ud-Ld)
In formula:LdWith UdRepresent search space minimum value and maximum respectively, d=1,2 ... D, D for solution dimension, due to
Vectorial solution is determined by five parameters, so D=5, quantity, nectar source scale NP and the maximum iteration of bee colony are according to actual conditions
Set.
Preferably, above-mentioned quantity, nectar source scale and maximum iteration to bee colony is respectively set to:Bee colony sum=
60, nectar source scale NP=30, maximum iteration=50.
The computational methods of fitness value are in above-mentioned steps 3:
fitiFor the fitness value in nectar source, fiFor object function solution be elliptical edge point matching degree, candidate ellipse picture
The set of vegetarian refreshments can be expressed as Xi=(s1,s2,…,sN), N is the quantity of all marginal points present on candidate's ellipse, target
The matching degree that function can be expressed as elliptical edge point is:
Xi(s1x,s1y) be defined as
(s in formula1x,s1y) be candidate it is oval in pixel point set S any pixel point coordinate.
Employ honeybee to scan for producing new nectar source in above-mentioned steps 4, calculate its adaptive value and preferable using greedy method selection
Nectar source specifically include:Search produces new nectar source near the nectar source of initialization, is shown below:
In formula:xidFor the nectar source i randomly generated in search space position, d is a random integers in [1, D],
Expression employs the selection of honeybee at random is one-dimensional to scan for;J ∈ { 1,2 ..., NP }, j ≠ i, xjdExpression is selected at random in NP nectar source
Select a nectar source for being not equal to i;It is the equally distributed random number in the scope of [- 1,1], determines perturbation amplitude, work as nectar source
vidFitness be better than xidWhen, using the method v of greediness selectionidInstead of xid, otherwise retain xid, that is, produce new candidate ellipse
Circle, Greedy principle is expressed as:
Wherein, f (vi) it is the new target function value for producing position, f (xi) be current nectar source target function value, when and only
When the target function value for newly producing position is smaller than the target function value of green molasses source position, the position in current nectar source is updated.
The selected new probability formula in each nectar source is in above-mentioned steps 5:
Honeybee select probability according to determined by above formula is followed, nectar source is selected according to roulette principle and essence is carried out in its vicinity
Fine searching, searching for formula is:
In formula:xidFor the nectar source i randomly generated in search space position, d is a random integers in [1, D],
Expression employs the selection of honeybee at random is one-dimensional to scan for.
Nectar source x in above-mentioned steps 6idBy limit iteration of threshold value without finding more preferable nectar source, then the nectar source will
It is abandoned, corresponding to employ the role of honeybee to be changed into investigating honeybee, investigation honeybee will randomly generate a new honey in search space
Source replaces xid, that is, candidate's ellipse is have updated, said process is represented by:
The position that optimal nectar source is remembered described in above-mentioned steps 7 is that each search formula obtains optimal solution i.e. most more than
Candidate close to Target ellipse is oval.
Compared with prior art, the beneficial effects of the present invention are:
1, artificial bee colony algorithm is used for the detection of leucocyte by the present invention, while preferable detection speed is kept, carries
The high precision of white blood cell detection.
2, the present invention has the advantages that detection speed is fast, Detection results are good, algorithm robustness is strong, preferably solve classics
Hough transform low memory and the problems such as operand is larger and particle cluster algorithm is easily trapped into locally optimal solution.
3, the present invention solves particle cluster algorithm and is easily trapped into locally optimal solution and other optimized algorithm accuracy of detection not
It is enough, the problems such as detection speed is unhappy.
Brief description of the drawings
Fig. 1 is the flow chart of the method for the white blood cell detection of the invention based on artificial bee colony algorithm.
Reference pictures of the Fig. 2 for the present invention applied to the cell of white blood cell detection.
Fig. 3 is the present invention to the result after leucocyte reference picture rim detection.
Fig. 4 is detection consequence of the particle cluster algorithm to leucocyte reference picture.
Fig. 5 is testing result of the present invention to leucocyte reference picture.
Embodiment
In order that the object, technical solution and advantage of invention are more clearly understood, it is right below in conjunction with accompanying drawing and case study on implementation
The present invention is further elaborated.It should be appreciated that specific implementation case described herein is only to explain the present invention,
It is not intended to limit the present invention.
As shown in figure 1, this method comprises the following steps:
Step 1:Using edge detection process image and extract all edge pixels and be stored in vector P
(rim detection is the primary bar of white blood cell detection to progress rim detection (as shown in Figure 2) to original image first
Part), the edge contour of leukocyte cell core ellipse in picture is extracted as shown in Figure 3, ellipses detection is then carried out, extraction institute
Some edge pixel storages are into vector P, P=(P1,P2,…,Pn), n is the summation of all edge pixels on image.
Step 2:Initialize population:Initialize nectar source xid, setting nectar source sum NP, control parameter limit, greatest iteration time
Number, t=1
The position in each nectar source represents the ellipse of a feasible solution, often produces a new candidate nectar source and represents
A kind of new candidate oval.Initialize the vector in all nectar sourcesOval fundamental equation can be with
It is expressed as:
ax2+bxy+cy2+ dx+ey+1=0
As five free parameters a, b, c, d, e, value meet b2Constituted during -4ac < 0 oval.So each is oval
It can be represented by five free parameters, in order to build oval (nectar source), arbitrarily extract five edge pixel points out in vector P
pi1,pi2,pi3,pi4,pi5(wherein any three points not point-blank), which is combined, can be formed into an ellipse, institute
So that each nectar source is encoded to an ellipse Xi.When carrying out ellipses detection, arbitrary five edge pixels in vector P
Five systems of linear equations on parameter that point is respectively constituted, are solved to it, if obtained solution meets condition, using people worker bee
Group's algorithm carries out selection optimization to solution.The purpose of the present invention be exactly processing is optimized to these vectors, obtain optimal solution to
Amount.The position of vector initialising candidate oval (nectar source):
xid=Ld+rand(0,1)(Ud-Ld)
In formula:LdWith UdSearch space minimum value and maximum, d=1,2 ... D are represented respectively.D is the dimension of solution, due to
Vectorial solution is determined by five parameters, so D=5.Quantity, nectar source scale and the maximum iteration of bee colony are set according to actual conditions
Put, including be set to bee colony sum=60, nectar source scale NP=30, maximum iteration=50.
Step 3:Calculate the fitness value in nectar source
The computational methods of fitness value are:
fitiFor the fitness value in nectar source, fiFor object function solution be elliptical edge point matching degree.The present invention is excellent
Change the quantity ratios of marginal points all in the quantity and vector P of the marginal point overlapped on obtained candidate's ellipse with vector P
Do the matching degree of elliptical edge point.Optimization refers to obtain best candidate ellipse (Target ellipse) from all feasible solutions.Such as
Fruit Target ellipse necessary being, in order to calculate the matching degree of elliptical edge point, we should first calculate the candidate actually obtained
The coordinate of oval edge pixel point.The set of candidate's ellipse pixel can be expressed as Xi=(s1,s2,…,sN), N is candidate
The quantity of all marginal points present on ellipse.The matching degree that object function can be expressed as elliptical edge point is:
Xi(s1x,s1y) be defined as
(s in formula1x,s1y) be candidate it is oval in pixel point set S any pixel point coordinate.The oval inspection of the present invention
The method of survey is the problem of belonging to maximum, it is therefore an objective to make fiValue it is maximum.Function fit is calculated according to fitnessi, in order to obtain
Maximal solution, the greedy less nectar source of back-and-forth method selection fitness, therefore new nectar source v is found in searchidAfterwards, according to fitness value
Size, in vidWith xidBetween using the greedy less nectar source of back-and-forth method selection fitness value, producing new nectar source, (candidate is ellipse
Circle).
Step 4:Employ honeybee to scan for producing new nectar source, calculate its adaptive value and preferable nectar source is selected using greedy method
Search produces new nectar source (new candidate is oval) formula near the nectar source (candidate is oval) of initialization:
In formula:xidFor the nectar source i randomly generated in search space position, d is a random integers in [1, D],
Expression employs the selection of honeybee at random is one-dimensional to scan for;J ∈ { 1,2 ..., NP }, j ≠ i, xjdExpression is selected at random in NP nectar source
Select a nectar source for being not equal to i;It is the equally distributed random number in the scope of [- 1,1], determines perturbation amplitude.Work as nectar source
vidFitness be better than xidWhen, using the method v of greediness selectionidInstead of xid, otherwise retain xid, that is, produce new candidate ellipse
Circle.Greedy principle is expressed as:
Wherein, f (vi) it is the new target function value for producing position, f (xi) be current nectar source target function value, when and only
When the target function value for newly producing position is smaller than the target function value of green molasses source position, the position in current nectar source is updated.
Step 5:Honeybee sharing information is employed, follows honeybee to calculate the selected probability in nectar source using roulette principle, while root
According to search formula, nearby the selected new probability formula in each nectar source of fine search is:
Honeybee select probability according to determined by listing is followed, nectar source is selected according to roulette principle and essence is carried out in its vicinity
Fine searching, searching for formula is:
In formula:xidFor the nectar source i randomly generated in search space position, d is a random integers in [1, D],
Expression employs the selection of honeybee at random is one-dimensional to scan for.
Step 6:Determine whether that food source continuous limit generations do not update, if then employing honeybee to be changed into investigating honeybee, detect
Look into the new nectar source of honeybee search generation and 8 are gone to step if not
Nectar source xidThreshold value limit is reached without finding more preferable nectar source by trial iterative search, then the nectar source
xidIt will be abandoned, it is corresponding to employ the role of honeybee to be changed into investigating honeybee.Investigation honeybee will randomly generate one in search space
New nectar source replaces xid, that is, have updated candidate oval.Said process is represented by:
Step 7:The best nectar source position of memory
The position in the optimal nectar source of memory, respectively searches for obtain optimal solution i.e. closest to Target ellipse with formula more than
Candidate it is oval.
Step:8:T=t+1;Judge whether to reach maximum iteration, output optimal solution is terminated if reaching, otherwise
Repeat step 3-7
The iterative algorithm set according to step 2, judges whether iterations reaches:If reaching iterations max=50,
Then stop iteration, obtain position and the target function value in optimal nectar source, the corresponding position in the nectar source is Target ellipse.If no
Iterations then repeat step 3-7 is reached, until reaching the maximum iteration set by step 2.
Step 9:Oval image is detected
Object function is higher, and explanation Ellipse Matching degree is higher, closer to Target ellipse, is detecting the result of leucocyte more just
Really.As Fig. 4 and Fig. 5 makes comparisons, it is evident that Fig. 5 testing result is more preferable.
Claims (10)
1. the detection method of the leucocyte based on artificial bee colony algorithm, it is characterised in that comprise the following steps:
Step 1:Using edge detection process image, and extract all edge pixels and be stored in vector P;
Step 2:Initialize population:Initialize nectar source xid, setting nectar source sum NP, control parameter limit, maximum iteration, t
=1;
Step 3:Calculate the fitness value in nectar source;
Step 4:Employ honeybee to scan for producing new nectar source, calculate its adaptive value and preferable nectar source is selected using greedy method;
Step 5:Honeybee sharing information is employed, follows honeybee to calculate the selected probability in nectar source using roulette principle, while according to searching
Rope formula nearby fine search;
Step 6:Determine whether that food source continuous limit generations do not update, if then employing honeybee to be changed into investigating honeybee, investigate honeybee
Search produces new nectar source;Step 8 is gone to if not;
Step 7:The optimal nectar source position of memory;
Step:8:T=t+1;Judge whether to reach maximum iteration, output optimal solution is terminated if reaching, is otherwise repeated
Step 3-7;
Step 9:The image of leucocyte is detected.
2. the detection method of the leucocyte as claimed in claim 1 based on artificial bee colony algorithm, it is characterised in that the step
Rim detection in 1 specifically includes the edge contour for extracting leukocyte cell core ellipse in picture, then carries out ellipses detection,
Extract all edge pixels to store into vector P, P=(P1,P2,…,Pn), n is the summation of all edge pixels on image.
3. the detection method of the leucocyte as claimed in claim 1 based on artificial bee colony algorithm, it is characterised in that the step
The position in the nectar source of each in 2 represents the ellipse of a feasible solution, often produces a new candidate nectar source and represents one
New candidate's ellipse is planted, the vector in all nectar sources is initializedOval fundamental equation can be represented
For:
ax2+bxy+cy2+ dx+ey+1=0
As five free parameters a, b, c, d, e, value meet b2Constituted during -4ac < 0 it is oval, so each ellipse can be by
Five free parameters are represented, in order to build ellipse, arbitrarily extract five edge pixel point p out in vector Pi1,pi2,pi3,pi4,
pi5, wherein any three points are not point-blank, an ellipse can be constituted by being combined, so each nectar source
It is encoded to an ellipse Xi, when carrying out ellipses detection, arbitrary five edge pixels point is respectively constituted in vector P five
On the system of linear equations of parameter, it is solved, if obtained solution meets condition, solution selected using artificial bee colony algorithm
Preferentially change.
4. the detection method of the leucocyte as claimed in claim 1 based on artificial bee colony algorithm, it is characterised in that the step
The position that nectar source is initialized in 2 is:
xid=Ld+rand(0,1)(Ud-Ld)
In formula:LdWith UdRepresent search space minimum value and maximum respectively, d=1,2 ... D, D for solution dimension, due to vector
Solution is determined by five parameters, so D=5, the quantity of bee colony, nectar source scale NP and iterations max are set according to actual conditions.
5. the detection method of the leucocyte as claimed in claim 4 based on artificial bee colony algorithm, it is characterised in that to bee colony
Quantity, nectar source scale and maximum iteration are respectively set to:Bee colony sum=60, nectar source scale NP=30, greatest iteration time
Number=50.
6. the detection method of the leucocyte as claimed in claim 1 based on artificial bee colony algorithm, it is characterised in that the step
The computational methods of fitness value are in 3:
fitiFor the fitness value in nectar source, fiFor object function solution be elliptical edge point matching degree, candidate ellipse pixel
Set can be expressed as Xi=(s1,s2,…,sN), N is the quantity of all marginal points present on candidate's ellipse, object function
The matching degree that elliptical edge point can be expressed as is:
Xi(s1x,s1y) be defined as
(s in formula1x,s1y) be candidate it is oval in pixel point set S any pixel point coordinate.
7. the detection method of the leucocyte as claimed in claim 1 based on artificial bee colony algorithm, it is characterised in that the step
Employ honeybee to scan for producing new nectar source in 4, calculate its adaptive value and select preferable nectar source to specifically include using greedy method:
Search produces new nectar source near the nectar source of initialization, is shown below:
In formula:xidFor the nectar source i randomly generated in search space position, d is a random integers in [1, D], is represented
Employ the selection of honeybee at random is one-dimensional to scan for;J ∈ { 1,2 ..., NP }, j ≠ i, xjdExpression randomly chooses one in NP nectar source
The individual nectar source for being not equal to i;It is the equally distributed random number in the scope of [- 1,1], perturbation amplitude is determined, as nectar source vid's
Fitness is better than xidWhen, using the method v of greediness selectionidInstead of xid, otherwise retain xid, that is, new candidate's ellipse is produced, is coveted
Greedy principle is expressed as:
Wherein, f (vi) it is the new target function value for producing position, f (xi) be current nectar source target function value, it is new that and if only if
Target function value hour of the target function value than green molasses source position of position is produced, the position in current nectar source is updated.
8. the detection method of the leucocyte as claimed in claim 1 based on artificial bee colony algorithm, it is characterised in that the step
The selected new probability formula in each nectar source is in 5:
Honeybee select probability according to determined by above formula is followed, nectar source is selected according to roulette principle and finely searched in its vicinity
Rope, searching for formula is:
In formula:xidFor the nectar source i randomly generated in search space position, d is a random integers in [1, D], is represented
Employ the selection of honeybee at random is one-dimensional to scan for.
9. the detection method of the leucocyte as claimed in claim 1 based on artificial bee colony algorithm, it is characterised in that the step
Nectar source x in 6idBy limit iteration of threshold value without finding more preferable nectar source, then the nectar source will be abandoned, and be corresponded to therewith
Employ honeybee role be changed into investigate honeybee, investigation honeybee replace x by a new nectar source is randomly generated in search spaceid, that is, update
Candidate is oval, and said process is represented by:
10. the detection method of the leucocyte as claimed in claim 1 based on artificial bee colony algorithm, its feature is in the step 7
Described in remember the position in optimal nectar source be that each search formula obtains optimal solution i.e. closest to the time of Target ellipse more than
Choosing is oval.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710169481.2A CN107045717A (en) | 2017-03-21 | 2017-03-21 | The detection method of leucocyte based on artificial bee colony algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710169481.2A CN107045717A (en) | 2017-03-21 | 2017-03-21 | The detection method of leucocyte based on artificial bee colony algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107045717A true CN107045717A (en) | 2017-08-15 |
Family
ID=59545601
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710169481.2A Pending CN107045717A (en) | 2017-03-21 | 2017-03-21 | The detection method of leucocyte based on artificial bee colony algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107045717A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108513234A (en) * | 2018-03-14 | 2018-09-07 | 南京邮电大学 | Based on the loud speaker volume optimization method for improving artificial bee colony algorithm |
CN109002923A (en) * | 2018-07-23 | 2018-12-14 | 宁波大学 | A kind of intercity multimode travel route planing method |
CN109919963A (en) * | 2019-03-14 | 2019-06-21 | 吉林大学 | A kind of vehicle paint method for detecting position of defect |
CN111276981A (en) * | 2020-05-08 | 2020-06-12 | 广东电网有限责任公司佛山供电局 | Voltage sag comprehensive treatment method for power grid side |
CN111985705A (en) * | 2020-08-13 | 2020-11-24 | 复旦大学 | Multipoint position shortest path calculation method |
CN112202538A (en) * | 2020-09-21 | 2021-01-08 | 齐齐哈尔大学 | Artificial bee colony OFDMA resource allocation method for guaranteeing fairness threshold value |
CN113066096A (en) * | 2021-04-06 | 2021-07-02 | 长沙证通云计算有限公司 | Edge detection method and system based on improved ant colony clustering algorithm |
CN113208569A (en) * | 2020-01-21 | 2021-08-06 | 复旦大学 | Pulse wave curve fitting method based on group algorithm |
CN114078174A (en) * | 2020-08-11 | 2022-02-22 | 复旦大学 | Pulse wave curve fitting method based on improved ABC algorithm |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101894273A (en) * | 2010-05-26 | 2010-11-24 | 北京航空航天大学 | Artificial bee colony refine edge potential field function-based unmanned plane target identification method |
CN104835151A (en) * | 2015-04-24 | 2015-08-12 | 南京邮电大学 | Improved artificial bee colony algorithm-based image registration method |
CN105956706A (en) * | 2016-05-09 | 2016-09-21 | 苏州泛能电力科技有限公司 | Distribution network reactive power optimization method based on improved artificial bee colony algorithm |
-
2017
- 2017-03-21 CN CN201710169481.2A patent/CN107045717A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101894273A (en) * | 2010-05-26 | 2010-11-24 | 北京航空航天大学 | Artificial bee colony refine edge potential field function-based unmanned plane target identification method |
CN104835151A (en) * | 2015-04-24 | 2015-08-12 | 南京邮电大学 | Improved artificial bee colony algorithm-based image registration method |
CN105956706A (en) * | 2016-05-09 | 2016-09-21 | 苏州泛能电力科技有限公司 | Distribution network reactive power optimization method based on improved artificial bee colony algorithm |
Non-Patent Citations (3)
Title |
---|
APRINALDI 等: "ArcPSO: Ellipse Detection Method using Particle Swarm Optimization andArc Combination", 《ICACSIS 2014》 * |
ERIK CUEVAS 等: "Multi-circle detection on images using artificial bee colony (ABC) optimization", 《SOFT COMPUTING》 * |
陈阿慧 等: "人工蜂群算法综述", 《智能计算机与应用》 * |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108513234A (en) * | 2018-03-14 | 2018-09-07 | 南京邮电大学 | Based on the loud speaker volume optimization method for improving artificial bee colony algorithm |
CN108513234B (en) * | 2018-03-14 | 2020-03-20 | 南京邮电大学 | Loudspeaker volume optimization method based on improved artificial bee colony algorithm |
CN109002923A (en) * | 2018-07-23 | 2018-12-14 | 宁波大学 | A kind of intercity multimode travel route planing method |
CN109919963A (en) * | 2019-03-14 | 2019-06-21 | 吉林大学 | A kind of vehicle paint method for detecting position of defect |
CN113208569A (en) * | 2020-01-21 | 2021-08-06 | 复旦大学 | Pulse wave curve fitting method based on group algorithm |
CN113208569B (en) * | 2020-01-21 | 2022-06-10 | 复旦大学 | Pulse wave curve fitting method based on group algorithm |
CN111276981B (en) * | 2020-05-08 | 2020-09-25 | 广东电网有限责任公司佛山供电局 | Voltage sag comprehensive treatment method for power grid side |
CN111276981A (en) * | 2020-05-08 | 2020-06-12 | 广东电网有限责任公司佛山供电局 | Voltage sag comprehensive treatment method for power grid side |
CN114078174A (en) * | 2020-08-11 | 2022-02-22 | 复旦大学 | Pulse wave curve fitting method based on improved ABC algorithm |
CN111985705A (en) * | 2020-08-13 | 2020-11-24 | 复旦大学 | Multipoint position shortest path calculation method |
CN111985705B (en) * | 2020-08-13 | 2024-06-28 | 复旦大学 | Multi-point position shortest path calculation method |
CN112202538A (en) * | 2020-09-21 | 2021-01-08 | 齐齐哈尔大学 | Artificial bee colony OFDMA resource allocation method for guaranteeing fairness threshold value |
CN112202538B (en) * | 2020-09-21 | 2023-08-22 | 齐齐哈尔大学 | Artificial bee colony OFDMA resource allocation method for guaranteeing fairness threshold |
CN113066096A (en) * | 2021-04-06 | 2021-07-02 | 长沙证通云计算有限公司 | Edge detection method and system based on improved ant colony clustering algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107045717A (en) | The detection method of leucocyte based on artificial bee colony algorithm | |
Li et al. | An improved ResNet based on the adjustable shortcut connections | |
GB2581758A (en) | Human hair style generation method based on multi-feature search and deformation | |
Xiao et al. | A fast method for particle picking in cryo-electron micrographs based on fast R-CNN | |
KR101687217B1 (en) | Robust face recognition pattern classifying method using interval type-2 rbf neural networks based on cencus transform method and system for executing the same | |
CN104112143A (en) | Weighted hyper-sphere support vector machine algorithm based image classification method | |
Ramdhani et al. | Hierarchical Decision Approach based on Neural Network and Genetic Algorithm method for single image classification of Pap smear | |
CN111524140B (en) | Medical image semantic segmentation method based on CNN and random forest method | |
CN111814845A (en) | Pedestrian re-identification method based on multi-branch flow fusion model | |
Wang et al. | Study on the method of transmission line foreign body detection based on deep learning | |
Dong et al. | An improved YOLOv5 network for lung nodule detection | |
CN111242971B (en) | Target tracking method based on improved double-center particle swarm optimization algorithm | |
CN112580445A (en) | Human body gait image visual angle conversion method based on generation of confrontation network | |
CN106845462A (en) | The face identification method of feature and cluster is selected while induction based on triple | |
CN104361135A (en) | Image retrieval method | |
Guehairia et al. | Deep random forest for facial age estimation based on face images | |
Yang et al. | Shape-based classification of environmental microorganisms | |
CN113836330A (en) | Image retrieval method and device based on generation antagonism automatic enhanced network | |
Budhi et al. | Acne segmentation and classification using region growing and self-organizing map | |
CN108446740B (en) | A kind of consistent Synergistic method of multilayer for brain image case history feature extraction | |
Jerandu et al. | Image Classification of Decapterus Macarellus Using Ridge Regression | |
Dong et al. | Dragon fruit disease image segmentation based on FCM algorithm and two-dimensional OTSU algorithm | |
CN105069445B (en) | A kind of face identification method based on improvement bacterial foraging algorithm | |
CN114299586A (en) | Intelligent deep learning system based on convolutional neural network | |
Cui et al. | Weighted particle swarm clustering algorithm for self-organizing maps |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170815 |
|
RJ01 | Rejection of invention patent application after publication |