CN105046712B - Based on the circle detection method that adaptability difference of Gaussian develops - Google Patents

Based on the circle detection method that adaptability difference of Gaussian develops Download PDF

Info

Publication number
CN105046712B
CN105046712B CN201510478231.8A CN201510478231A CN105046712B CN 105046712 B CN105046712 B CN 105046712B CN 201510478231 A CN201510478231 A CN 201510478231A CN 105046712 B CN105046712 B CN 105046712B
Authority
CN
China
Prior art keywords
individual
edge pixel
population
value
goes
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.)
Expired - Fee Related
Application number
CN201510478231.8A
Other languages
Chinese (zh)
Other versions
CN105046712A (en
Inventor
郭肇禄
岳雪芝
尹宝勇
杨火根
鄢化彪
刘松华
吴志健
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangxi University of Science and Technology
Original Assignee
Jiangxi University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangxi University of Science and Technology filed Critical Jiangxi University of Science and Technology
Priority to CN201510478231.8A priority Critical patent/CN105046712B/en
Publication of CN105046712A publication Critical patent/CN105046712A/en
Application granted granted Critical
Publication of CN105046712B publication Critical patent/CN105046712B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a kind of circle detection method developed based on adaptability difference of Gaussian, the present invention devises two kinds of Gaussian mutation strategies that global search guiding and Local Search are oriented to, according to current Evolution States suitable Gaussian mutation strategy is adaptively selected from both Gaussian mutation strategies in the evolutionary process of loop truss to produce variation individual, and adaptively adjustment crossover probability generates new individual, coordinates the balance between convergence rate and population diversity；Above-mentioned steps are repeated until meeting end condition, by the optimum individual obtained in detection process, you can obtain the central coordinate of circle and radius of circle in digital image；The present invention can accelerate the speed of loop truss, improve the efficiency and precision of loop truss.

Description

Based on the circle detection method that adaptability difference of Gaussian develops
Technical field
The present invention relates to digital image processing field, more particularly, to a kind of circle inspection developed based on adaptability difference of Gaussian Survey method.
Background technology
The detection of circle is circular object identification and inspection in the fields such as industrial automation, robot vision in Digital Image Processing The base support technology of survey.Traditional circle detection method has the methods such as Hough transform, least square method, gradient.But these are passed System method has that operand is larger, and detection speed is slower, and real-time is not strong, when especially processing the more complicated image of background, Accuracy of detection shortcoming not high.
Round test problems can sum up as an optimization problem.Differential Evolution Algorithm is a kind of solving-optimizing problem Effective ways, it has been successfully applied in many engineering practices, and obtains gratifying solving result.For example, Wang Ling etc. has invented a kind of optimal deployment side of large-scale industry wireless sensor network based on differential evolution algorithm in 2010 Method (the patent No.：201010290702.X), Liu Zhigang and Zeng Xueqiang invented a kind of based on improved differential evolution calculation in 2011 Method for Reactive Power Optimization in Power (the patent No. of method：201110130062.0), Li Ni etc. invented one kind and has been based in 2012 Airfoil Optimization method (the patent No. of the parallel differential evolution algorithm of OpenCL：201210208326.4).Although difference is drilled Change algorithm and certain success is obtained in many engineering fields, but conventional differential evolution algorithmic carries out loop truss to digital picture When, easily occur being absorbed in local optimum and the slow shortcoming of convergence rate, when especially processing complex background image, its accuracy of detection Still await improving.
The content of the invention
The present invention mainly solves the technical problem existing for prior art；For conventional differential evolution algorithmic to digitized map As when carrying out loop truss, easily occurring being absorbed in local optimum and the slow shortcoming of convergence rate, and during treatment complex background image, Accuracy of detection shortcoming not high, proposes a kind of circle detection method developed based on adaptability difference of Gaussian.The present invention can accelerate The speed of loop truss, improves the precision of loop truss, and avoids being absorbed in local optimum to a certain extent.
Technical scheme：A kind of circle detection method developed based on adaptability difference of Gaussian, is comprised the following steps：
Step 1, piece image IM is gathered using digital image sensor, then carries out rim detection to image IM, then right The image that rim detection is obtained performs binarization operation and obtains bianry image BI, and wherein pixel value is 1 expression edge pixel, pixel It is worth for 0 represents non-edge pixels；
Step 2, to bianry image BI successively by from left to right, all edge pixels are carried out layout by order from top to bottom Sequence number, then according to the edge pixel sequence number finished successively their two-dimensional coordinate storage to edge pixel list of coordinates In EList, wherein EList={ E1,E2,...,Ek,...,EEN, subscript k=1 ..., EN, wherein EN is edge pixel sum, Ek=[Ek,1Ek,2] it is two-dimensional coordinate value of k-th edge pixel in bianry image BI, Ek,1It is k-th edge pixel two Lateral coordinates value in value image BI, Ek,2It is longitudinal coordinate value of k-th edge pixel in bianry image BI；
Step 3, user's initiation parameter, the initiation parameter includes Population Size Popsize, and maximum evaluates number of times MAX_FEs；
Step 4, current evolution algebraically t=0, Evaluation: Current number of times FEs=0；
Step 5, randomly generates initial populationWherein：Individual subscript i= 1 ..., Popsize, andIt is population PtIn i-th it is individual, its random initializtion formula is：
Wherein dimension subscript j=1,2,3；It is in population PtIn i-th it is individual, store 3 spans and arrived 1 Real number between EN, rand (0,1) is to obey equally distributed random real number between [0,1] to produce function；
Step 6, initial population PtIn each individual hybrid rateThen population P is randomly generatedtIn per each and every one The Gaussian mutation strategy sequence number of bodyWherein individual subscript i=1 ..., Popsize, RandInt () is that random integers produce function, % to be accorded with for complementation；
Step 7, calculates population PtIn each is individualAdaptive valueWherein individual subscript i=1 ..., Popsize, calculates individualAdaptive valueComprise the following steps that：
Step 7.1, obtains individualThree real numbers of middle storageFour houses are carried out to these three real numbers again Five enter to round, and respectively obtain three edge pixel sequence numbers:Then in edge pixel list of coordinates EList In obtain the two-dimensional coordinate value of the edge pixel that these three edge pixel sequence numbers are represented respectively：
Step 7.2, judges these three two-dimensional coordinate valuesWhether on same straight line, if it is Order is individualAdaptive value be 1.0, then terminate individualityAdaptive value calculating process；Otherwise go to step 7.3；
Step 7.3, by three two-dimensional coordinate valuesDetermine a circular path CP, and use Strokes and dots circule method calculates the coordinate of each point on circular path CP, and it is CPN to write down the sum put on circular path CP；
Step 7.4, makes counter EPN=0；
Step 7.5, judges that the corresponding pixel in bianry image BI of the coordinate of each point on circular path CP is successively No is edge pixel, if it is makes EPN=EPN+1；Otherwise keep the value of EPN constant；
Step 7.6, order is individualAdaptive value beWherein adaptive value is smaller, shows individual more outstanding；
Step 8, Evaluation: Current number of times FEs=FEs+Popsize；
Step 9, preserves population PtThe minimum individuality of middle adaptive value is optimum individual Bestt；
Step 10, makes counter i=1；
Step 11, if counter i is more than Population Size Popsize, goes to step 18, otherwise goes to step 12；
Step 12, calculates individualCurrent hybrid rateComputing formula is as follows:
Wherein r1 is the real number randomly generated between [0,1]；
Step 13, according to individualityGaussian mutation strategy sequence numberProduce an experiment individualSpecific steps It is as follows:
Step 13.1, makes counter j=1；
Step 13.2, randomly generates a positive integer jRand between [1, D]；
Step 13.3, randomly generates two unequal positive integer RI1, RI2 between [1, Popsize]；
Step 13.4, if counter j is more than D, goes to step 13.16, otherwise goes to step 13.5；
Step 13.5, produces a random real number r2 between [0,1], if r2 is less than individualityCurrent hybrid rateOr jRand is equal to counter j, then go to step 13.6, otherwise goes to step 13.14；
Step 13.6, ifStep 13.7 is gone to equal to 1 and performs the Gaussian mutation strategy that Local Search is oriented to, Otherwise go to step 13.10 and perform the Gaussian mutation strategy that global search is oriented to；
Step 13.7, average
Step 13.8, variance
Step 13.9,Wherein NormalRand (Mval, SDVal) is represented With Mval as average, SDVal produces function for the Gauss number of variance, then goes to step 13.15；
Step 13.10, random weights RW1=rand (0,1)；
Step 13.11, average
Step 13.12, variance
Step 13.13,Go to step 13.15；
Step 13.14,
Step 13.15, makes counter j=j+1, goes to step 13.4；
Step 13.16, calculates experiment individualAdaptive valueGo to step 14；
Step 14, as follows in individualityIt is individual with experimentBetween select excellent individual and enter of future generation kind Group：
Step 15, as follows more new individualGaussian mutation strategy sequence numberAnd hybrid rate
Step 16, makes counter i=i+1；
Step 17, goes to step 11；
Step 18, Evaluation: Current number of times FEs=FEs+Popsize preserves population PtThe minimum individuality of middle adaptive value is for most Excellent individual Bestt；
Step 19, current evolution algebraically t=t+1；
Step 20, repeat step 10 to step 19 terminates after Evaluation: Current number of times FEs reaches MAX_FEs, from execution The optimum individual Best that process is obtainedtThree real numbers of middle acquisitionAgain these three real numbers are carried out with four houses five Enter to round, respectively obtain three edge pixel sequence numbers:Then in edge pixel list of coordinates EList In obtain the two-dimensional coordinate value of the edge pixel that these three edge pixel sequence numbers are represented respectively：Pass through These three two-dimensional coordinate values can obtain the circle for detecting.
The present invention devises two kinds of Gaussian mutation strategies that global search guiding and Local Search are oriented to, in drilling for loop truss Suitable Gauss is adaptively selected from both Gaussian mutation strategies during change according to current Evolution States to become Different strategy produces variation individual, and adaptively adjustment crossover probability generates new individual, coordinates convergence rate many with population Balance between sample, from the speed for accelerating loop truss, improves the precision of loop truss, and reduction is absorbed in the probability of local optimum；With Congenic method is compared, and the present invention can accelerate the speed of loop truss, improves the precision of loop truss, and avoids falling into a certain extent Enter local optimum.
Brief description of the drawings
Fig. 1 is loop truss image in embodiment.
Fig. 2 is the edge result of loop truss image.
Fig. 3 is the result that loop truss is carried out using the present invention.
Specific embodiment
Below by embodiment, and with reference to accompanying drawing, technical scheme is described in further detail.
Embodiment：
The present embodiment is based on Fig. 1 loop truss images, and specific implementation step of the invention is as follows：
Step 1, piece image IM (as shown in Figure 1) is gathered using digital image sensor, then carries out side to image IM Edge is detected, then the image obtained to rim detection performs binarization operation and obtains bianry image BI (as shown in Figure 2), wherein pixel It is worth for 1 represents edge pixel, pixel value is 0 expression non-edge pixels；
Step 2, to bianry image BI successively by from left to right, all edge pixels are carried out layout by order from top to bottom Sequence number, then according to the edge pixel sequence number finished successively their two-dimensional coordinate storage to edge pixel list of coordinates In EList, wherein EList={ E1,E2,...,Ek,...,EEN, subscript k=1 ..., EN, wherein EN is edge pixel sum, Ek=[Ek,1Ek,2] it is two-dimensional coordinate value of k-th edge pixel in bianry image BI, Ek,1It is k-th edge pixel two Lateral coordinates value in value image BI, Ek,2It is longitudinal coordinate value of k-th edge pixel in bianry image BI；
Step 3, user's initiation parameter, the initiation parameter includes Population Size Popsize=20, and maximum is evaluated secondary Number MAX_FEs=1000；
Step 4, current evolution algebraically t=0, Evaluation: Current number of times FEs=0；
Step 5, randomly generates initial populationWherein：Individual subscript i= 1 ..., Popsize, andIt is population PtIn i-th it is individual, its random initializtion formula is：
Wherein dimension subscript j=1,2,3；It is in population PtIn i-th it is individual, store 3 spans and arrived 1 Real number between EN, rand (0,1) is to obey equally distributed random real number between [0,1] to produce function；
Step 6, initial population PtIn each individual hybrid rateThen population P is randomly generatedtIn per each and every one The Gaussian mutation strategy sequence number of bodyWherein individual subscript i=1 ..., Popsize, RandInt () is that random integers produce function, % to be accorded with for complementation；
Step 7, calculates population PtIn each is individualAdaptive valueWherein individual subscript i=1 ..., Popsize, calculates individualAdaptive valueComprise the following steps that：
Step 7.1, obtains individualThree real numbers of middle storageFour houses are carried out to these three real numbers again Five enter to round, and respectively obtain three edge pixel sequence numbers:Then in edge pixel list of coordinates EList In obtain the two-dimensional coordinate value of the edge pixel that these three edge pixel sequence numbers are represented respectively：
Step 7.2, judges these three two-dimensional coordinate valuesWhether on same straight line, if it is Order is individualAdaptive value be 1.0, then terminate individualityAdaptive value calculating process；Otherwise go to step 7.3；
Step 7.3, by three two-dimensional coordinate valuesDetermine a circular path CP, and use Strokes and dots circule method calculates the coordinate of each point on circular path CP, and it is CPN to write down the sum put on circular path CP；
Step 7.4, makes counter EPN=0；
Step 7.5, judges that the corresponding pixel in bianry image BI of the coordinate of each point on circular path CP is successively No is edge pixel, if it is makes EPN=EPN+1；Otherwise keep the value of EPN constant；
Step 7.6, order is individualAdaptive value beWherein adaptive value is smaller, shows individual more outstanding；
Step 8, Evaluation: Current number of times FEs=FEs+Popsize；
Step 9, preserves population PtThe minimum individuality of middle adaptive value is optimum individual Bestt；
Step 10, makes counter i=1；
Step 11, if counter i is more than Population Size Popsize, goes to step 18, otherwise goes to step 12；
Step 12, calculates individualCurrent hybrid rateComputing formula is as follows:
Wherein r1 is the real number randomly generated between [0,1]；
Step 13, according to individualityGaussian mutation strategy sequence numberProduce an experiment individualSpecific steps It is as follows:
Step 13.1, makes counter j=1；
Step 13.2, randomly generates a positive integer jRand between [1, D]；
Step 13.3, randomly generates two unequal positive integer RI1, RI2 between [1, Popsize]；
Step 13.4, if counter j is more than D, goes to step 13.16, otherwise goes to step 13.5；
Step 13.5, produces a random real number r2 between [0,1], if r2 is less than individualityCurrent hybrid rateOr jRand is equal to counter j, then go to step 13.6, otherwise goes to step 13.14；
Step 13.6, ifStep 13.7 is gone to equal to 1 and performs the Gaussian mutation strategy that Local Search is oriented to, Otherwise go to step 13.10 and perform the Gaussian mutation strategy that global search is oriented to；
Step 13.7, average
Step 13.8, variance
Step 13.9,Wherein NormalRand (Mval, SDVal) is represented With Mval as average, SDVal produces function for the Gauss number of variance, then goes to step 13.15；
Step 13.10, random weights RW1=rand (0,1)；
Step 13.11, average
Step 13.12, variance
Step 13.13,Go to step 13.15；
Step 13.14,
Step 13.15, makes counter j=j+1, goes to step 13.4；
Step 13.16, calculates experiment individualAdaptive valueGo to step 14；
Step 14, as follows in individualityIt is individual with experimentBetween select excellent individual and enter of future generation kind Group：
Step 15, as follows more new individualGaussian mutation strategy sequence numberAnd hybrid rate
Step 16, makes counter i=i+1；
Step 17, goes to step 11；
Step 18, Evaluation: Current number of times FEs=FEs+Popsize preserves population PtThe minimum individuality of middle adaptive value is for most Excellent individual Bestt；
Step 19, current evolution algebraically t=t+1；
Step 20, repeat step 10 to step 19 terminates after Evaluation: Current number of times FEs reaches MAX_FEs, from execution The optimum individual Best that process is obtainedtThree real numbers of middle acquisitionFour houses are carried out to these three real numbers again Five enter to round, and respectively obtain three edge pixel sequence numbers:Then in edge pixel list of coordinates The two-dimensional coordinate value of the edge pixel that these three edge pixel sequence numbers are represented respectively is obtained in EList：The circle for detecting can be obtained by these three two-dimensional coordinate values, as shown in Figure 3.
Specific embodiment described herein is only to the spiritual explanation for example of the present invention.Technology neck belonging to of the invention The technical staff in domain can be made various modifications or supplement to described specific embodiment or be replaced using similar mode Generation, but without departing from spirit of the invention or surmount scope defined in appended claims.

Claims (1)

1. it is a kind of based on adaptability difference of Gaussian develop circle detection method, it is characterised in that comprise the following steps：
Step 1, piece image IM is gathered using digital image sensor, then carries out rim detection to image IM, then to edge The image that detection is obtained performs binarization operation and obtains bianry image BI, and wherein pixel value is 1 expression edge pixel, and pixel value is 0 represents non-edge pixels；
Step 2, to bianry image BI successively by from left to right, all edge pixels are carried out layout sequence by order from top to bottom Number, then according to the edge pixel sequence number finished successively their two-dimensional coordinate storage to edge pixel list of coordinates EList In, wherein EList={ E1,E2,...,Ek,...,EEN, subscript k=1 ..., EN, wherein EN is edge pixel sum, Ek= [Ek,1Ek,2] it is two-dimensional coordinate value of k-th edge pixel in bianry image BI, Ek,1It is k-th edge pixel in binary map As the lateral coordinates value in BI, Ek,2It is longitudinal coordinate value of k-th edge pixel in bianry image BI；
Step 3, user's initiation parameter, the initiation parameter includes Population Size Popsize, and maximum evaluates number of times MAX_ FEs；
Step 4, current evolution algebraically t=0, Evaluation: Current number of times FEs=0；
Step 5, randomly generates initial populationWherein：Individual subscript i=1 ..., Popsize, andIt is population PtIn i-th it is individual, its random initializtion formula is：
B i , j t = 1 + r a n d ( 0 , 1 ) × ( E N - 1 )
Wherein dimension subscript j=1,2,3；It is in population PtIn i-th it is individual, store 3 spans 1 to EN it Between real number, rand (0,1) is to obey equally distributed random real number between [0,1] to produce function；
Step 6, initial population PtIn each individual hybrid rateThen population P is randomly generatedtIn each individuality Gaussian mutation strategy sequence numberWherein individuality subscript i=1 ..., Popsize, randInt () For random integers produce function, % to be accorded with for complementation；
Step 7, calculates population PtIn each is individualAdaptive valueWherein individual subscript i=1 ..., Popsize, Calculate individualAdaptive valueComprise the following steps that：
Step 7.1, obtains individualThree real numbers of middle storageAgain these three real numbers round up taking It is whole, respectively obtain three edge pixel sequence numbers:Then obtained in edge pixel list of coordinates EList The two-dimensional coordinate value of the edge pixel that these three edge pixel sequence numbers are represented respectively：
Step 7.2, judges these three two-dimensional coordinate valuesWhether on same straight line, if it is make individual BodyAdaptive value be 1.0, then terminate individualityAdaptive value calculating process；Otherwise go to step 7.3；
Step 7.3, by three two-dimensional coordinate valuesDetermine strokes and dots circle in a circular path CP, and use Method calculates the coordinate of each point on circular path CP, and it is CPN to write down the sum put on circular path CP；
Step 7.4, makes counter EPN=0；
Step 7.5, judge successively on circular path CP the corresponding pixel in bianry image BI of the coordinate of each point whether be Edge pixel, if it is makes EPN=EPN+1；Otherwise keep the value of EPN constant；
Step 7.6, order is individualAdaptive value beWherein adaptive value is smaller, shows individual more outstanding；
Step 8, Evaluation: Current number of times FEs=FEs+Popsize；
Step 9, preserves population PtThe minimum individuality of middle adaptive value is optimum individual Bestt；
Step 10, makes counter i=1；
Step 11, if counter i is more than Population Size Popsize, goes to step 18, otherwise goes to step 12；
Step 12, calculates individualCurrent hybrid rateComputing formula is as follows:
Wherein r1 is the real number randomly generated between [0,1]；
Step 13, according to individualityGaussian mutation strategy sequence numberProduce an experiment individualSpecific steps are such as Under:
Step 13.1, makes counter j=1；
Step 13.2, randomly generates a positive integer jRand between [1, D]；
Step 13.3, randomly generates two unequal positive integer RI1, RI2 between [1, Popsize]；
Step 13.4, if counter j is more than D, goes to step 13.16, otherwise goes to step 13.5；
Step 13.5, produces a random real number r2 between [0,1], if r2 is less than individualityCurrent hybrid rate Or jRand is equal to counter j, then go to step 13.6, otherwise goes to step 13.14；
Step 13.6, ifStep 13.7 is gone to equal to 1 and performs the Gaussian mutation strategy that Local Search is oriented to, otherwise Go to step 13.10 and perform the Gaussian mutation strategy that global search is oriented to；
Step 13.7, average
Step 13.8, variance
Step 13.9,Wherein NormalRand (Mval, SDVal) is represented with Mval It is average, SDVal produces function, then goes to step 13.15 for the Gauss number of variance；
Step 13.10, random weights RW1=rand (0,1)；
Step 13.11, average
Step 13.12, variance
Step 13.13,Go to step 13.15；
Step 13.14,
Step 13.15, makes counter j=j+1, goes to step 13.4；
Step 13.16, calculates experiment individualAdaptive valueGo to step 14；
Step 14, as follows in individualityIt is individual with experimentBetween select excellent individual and enter population of future generation：
Step 15, as follows more new individualGaussian mutation strategy sequence numberAnd hybrid rate
Step 16, makes counter i=i+1；
Step 17, goes to step 11；
Step 18, Evaluation: Current number of times FEs=FEs+Popsize preserves population PtThe minimum individuality of middle adaptive value is optimum individual Bestt；
Step 19, current evolution algebraically t=t+1；
Step 20, repeat step 10 to step 19 terminates after Evaluation: Current number of times FEs reaches MAX_FEs, from implementation procedure The optimum individual Best for obtainingtThree real numbers of middle acquisitionThese three real numbers are rounded up again Round, respectively obtain three edge pixel sequence numbers:Then in edge pixel list of coordinates EList In obtain the two-dimensional coordinate value of the edge pixel that these three edge pixel sequence numbers are represented respectively：Pass through These three two-dimensional coordinate values can obtain the circle for detecting.
CN201510478231.8A 2015-08-07 2015-08-07 Based on the circle detection method that adaptability difference of Gaussian develops Expired - Fee Related CN105046712B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510478231.8A CN105046712B (en) 2015-08-07 2015-08-07 Based on the circle detection method that adaptability difference of Gaussian develops

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510478231.8A CN105046712B (en) 2015-08-07 2015-08-07 Based on the circle detection method that adaptability difference of Gaussian develops

Publications (2)

Publication Number Publication Date
CN105046712A CN105046712A (en) 2015-11-11
CN105046712B true CN105046712B (en) 2017-06-30

Family

ID=54453230

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510478231.8A Expired - Fee Related CN105046712B (en) 2015-08-07 2015-08-07 Based on the circle detection method that adaptability difference of Gaussian develops

Country Status (1)

Country Link
CN (1) CN105046712B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106023164B (en) * 2016-05-12 2018-07-10 江西理工大学 Using the ellipse detection method of mixing harmonic search algorithm
CN106339573B (en) * 2016-07-24 2018-08-17 江西理工大学 The rare-earth mining area underground water total nitrogen concentration flexible measurement method of artificial bee colony optimization
CN108665451B (en) * 2018-05-04 2022-02-25 江西理工大学 Circle detection method based on ternary Gaussian difference evolution algorithm
CN108595871B (en) * 2018-05-04 2022-07-01 江西理工大学 Hydraulic support optimal design method based on mean difference evolution
CN112381849B (en) * 2020-11-12 2022-08-02 江西理工大学 Image edge detection method based on adaptive differential evolution

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102750522A (en) * 2012-06-18 2012-10-24 吉林大学 Method for tracking targets
CN102819652A (en) * 2012-08-22 2012-12-12 武汉大学 Mechanical parameter optimization design method based on adaptive reverse differential evolution
CN103684352A (en) * 2013-12-18 2014-03-26 中国电子科技集团公司第五十四研究所 Particle filtering method based on differential evolution
CN103902783A (en) * 2014-04-11 2014-07-02 北京工业大学 Draining pipeline network optimizing method based on generalized reverse learning difference algorithm
CN104318575A (en) * 2014-11-04 2015-01-28 江西理工大学 Multi-threshold image segmentation method based on comprehensive learning differential evolution algorithm
CN104462759A (en) * 2014-11-04 2015-03-25 江西理工大学 Water quality model parameter identification method based on reverse simplification differential evolution algorithm
CN104715490A (en) * 2015-04-09 2015-06-17 江西理工大学 Navel orange image segmenting method based on adaptive step size harmony search algorithm
CN104715124A (en) * 2015-04-09 2015-06-17 江西理工大学 Truss size optimization design method based on cloud model differential evolution algorithm

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7187800B2 (en) * 2002-08-02 2007-03-06 Computerized Medical Systems, Inc. Method and apparatus for image segmentation using Jensen-Shannon divergence and Jensen-Renyi divergence

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102750522A (en) * 2012-06-18 2012-10-24 吉林大学 Method for tracking targets
CN102819652A (en) * 2012-08-22 2012-12-12 武汉大学 Mechanical parameter optimization design method based on adaptive reverse differential evolution
CN103684352A (en) * 2013-12-18 2014-03-26 中国电子科技集团公司第五十四研究所 Particle filtering method based on differential evolution
CN103902783A (en) * 2014-04-11 2014-07-02 北京工业大学 Draining pipeline network optimizing method based on generalized reverse learning difference algorithm
CN104318575A (en) * 2014-11-04 2015-01-28 江西理工大学 Multi-threshold image segmentation method based on comprehensive learning differential evolution algorithm
CN104462759A (en) * 2014-11-04 2015-03-25 江西理工大学 Water quality model parameter identification method based on reverse simplification differential evolution algorithm
CN104715490A (en) * 2015-04-09 2015-06-17 江西理工大学 Navel orange image segmenting method based on adaptive step size harmony search algorithm
CN104715124A (en) * 2015-04-09 2015-06-17 江西理工大学 Truss size optimization design method based on cloud model differential evolution algorithm

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
"Road Pavement Crack Automatic Detection by MMS Images";A. Mancini等;《Control & Automation (MED)》;20130628;第1589-1596页 *
"差分进化算法研究进展";汪慎文等;《武汉大学学报（理学版）》;20140824;第60卷(第4期);第283-292页 *

Also Published As

Publication number Publication date
CN105046712A (en) 2015-11-11

Similar Documents

Publication Publication Date Title
CN105046712B (en) Based on the circle detection method that adaptability difference of Gaussian develops
Al-Jarrah et al. A novel edge detection algorithm for mobile robot path planning
CN102346850B (en) DataMatrix bar code area positioning method under complex metal background
Wang et al. A hybrid flower pollination algorithm based modified randomized location for multi-threshold medical image segmentation
CN104816306A (en) Robot, robot system, control device and control method
CN108279692A (en) A kind of UUV dynamic programming methods based on LSTM-RNN
CN105528638A (en) Method for grey correlation analysis method to determine number of hidden layer characteristic graphs of convolutional neural network
CN104899892B (en) A kind of quickly star map image asterism extracting method
CN110111430A (en) One kind extracting quadric method from three-dimensional point cloud
CN110309854A (en) A kind of signal modulation mode recognition methods and device
Cohen et al. Crater detection via genetic search methods to reduce image features
CN109903282A (en) A kind of method for cell count, system, device and storage medium
Vallet et al. A multi-label convolutional neural network for automatic image annotation
CN110751183A (en) Image data classification model generation method, image data classification method and device
Gilles et al. Metagraspnet: A large-scale benchmark dataset for scene-aware ambidextrous bin picking via physics-based metaverse synthesis
CN107123138B (en) Based on vanilla-R point to the point cloud registration method for rejecting strategy
CN104239411B (en) A kind of detection method of the lattice-shaped radar based on color, position cluster and Corner Detection
CN103279960A (en) Human body hidden thing image segmentation method based on X-ray back scattering image
Djemame et al. Combining cellular automata and particle swarm optimization for edge detection
CN104156944A (en) Change detection method based on NSGA-II evolutionary algorithm
Vasavada et al. A Hybrid method for detection of edges in grayscale images
Liu et al. Deep learning for picking point detection in dense cluster
CN108665451B (en) Circle detection method based on ternary Gaussian difference evolution algorithm
Zailan et al. YOLO-based Network Fusion for Riverine Floating Debris Monitoring System
CN104036234A (en) Image identification method for crater

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20170630

CF01 Termination of patent right due to non-payment of annual fee