CN106447673A - Chip pin extraction method under non-uniform illumination condition - Google Patents

Chip pin extraction method under non-uniform illumination condition Download PDF

Info

Publication number
CN106447673A
CN106447673A CN201610872992.6A CN201610872992A CN106447673A CN 106447673 A CN106447673 A CN 106447673A CN 201610872992 A CN201610872992 A CN 201610872992A CN 106447673 A CN106447673 A CN 106447673A
Authority
CN
China
Prior art keywords
pin
image
value
connected domain
chip pin
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
Application number
CN201610872992.6A
Other languages
Chinese (zh)
Inventor
高会军
靳万鑫
滕军
于金泳
杨宪强
林伟阳
孙光辉
李湛
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.)
Harbin Institute of Technology
Original Assignee
Harbin Institute of 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 Harbin Institute of Technology filed Critical Harbin Institute of Technology
Priority to CN201610872992.6A priority Critical patent/CN106447673A/en
Publication of CN106447673A publication Critical patent/CN106447673A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G06T2207/20221—Image fusion; Image merging
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30148—Semiconductor; IC; Wafer

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

The invention relates to a chip pin extraction method under a non-uniform illumination condition. Under an existing non-uniform illumination condition, by using a traditional chip pin extraction technology, an extraction effect is poor, complexity is high and robustness is poor. By using the method in the invention, the above problems are solved. The method comprises the following steps of 1, acquiring an original gray level image; 2, using a dynamic threshold segmentation method to carry out crude extraction of a chip pin and acquiring a two-value pin image; 3, acquiring a binary pin image which is screened via a connected domain; 4, calculating a center position of each effective connected domain in the binary pin image; calculating a spacing between each adjacent pins and taking a minimum value in the spacing as a pin spacing typical value; 5, acquiring a pin local binary image; 6, acquiring a pin mask image; and 7, carrying out AND operation on the pin mask image acquired from the step 6 and the original gray level image and completing extraction of the chip pin. The method is used in the pin extraction field.

Description

A kind of chip pin extracting method under the conditions of inhomogeneous illumination
Technical field
The present invention relates to the chip pin extracting method under the conditions of inhomogeneous illumination.
Background technology
High-performance chip mounter is one of the nucleus equipment in surface installation technique (SMT) field, can complete at a high speed, in high precision Component mounter.Vision detection system in chip mounter is prevented from component pin because deforming or lacking to the detection of chip pin And the paster mistake for causing, and chip pin extracts the committed step as chip pin detection, directly determines final detection As a result precision and reliability.
Existing chip pin extracting method is mainly based upon image Segmentation Technology.Wherein, global thresholdization segmentation be Simple effective method the most in many dividing methods, it be using the target to be extracted in image with image background in gray scale spy Property on difference and image is regarded as two class regions with different grey-scale, choose the rational threshold value of comparison, you can real The extraction of target in existing image.However, under actual application conditions, the restriction of working condition causes industrial camera in collection figure As when the change of illumination occurs, cause the image for obtaining the uneven situation of Luminance Distribution occur.In this case, simply Global threshold dividing method is no longer suitable for.
For the problems referred to above, the existing and actual solution for adopting is the method for using Dynamic global threshold segmentation instead, However, this kind of method is easy to produce over-segmentation or less divided so that final extraction accuracy reduces.In addition, existing Person also proposed a series of image segmentation algorithms based on threshold value：Maximum entropy threshold split-run, iterative threshold segmentation method, minimum are by mistake Difference thresholding method and maximum variance between clusters (Otsu method) etc..However, the most of complexity height of these algorithms, it is impossible to be applied to The requirement of quick detection in commercial production.Maximum variance between clusters (Otsu method) therein although its algorithm is simple, and with good Good segmentation performance, but the algorithm limits its use to the sensitivity of the interference factors such as noise.
When the present invention carries out chip pin extraction for existing image segmentation algorithm under the conditions of inhomogeneous illumination, extract Effect is poor, complexity is high, the problem of poor robustness, proposes a kind of local threshold that dynamic thresholding method and Otsu algorithm combine Partitioning algorithm, the algorithm can improve the efficiency of chip pin extraction and precision.
Content of the invention
The invention aims under the conditions of solving existing inhomogeneous illumination, traditional chip pin extractive technique is present Extraction effect is poor, complexity is high, the problem of poor robustness, and proposes the chip pin extracting method under the conditions of inhomogeneous illumination.
A kind of chip pin extracting method detailed process under the conditions of inhomogeneous illumination is：
Step one, the industrial component image to industrial camera collection carry out gray processing process, obtain original-gray image；
Step 2, the original-gray image obtained by step one, carry out chip pin using dynamic threshold segmentation method Thick extraction, obtains two-value pinoutss picture；
Step 3, the two-value pinoutss picture to obtaining in step 2 carry out connected component labeling, and to marked connected domain Screened, retained the connected domain of the number more than predetermined threshold value T of contained pixel in connected domain, obtained is screened through connected domain Two-value pinoutss picture afterwards；
Each in step 4, the two-value pinoutss picture after connected domain screening obtained by step 3 is effectively connected Domain, asks for its center；The spacing of each adjacent leads is calculated, using the minima in spacing as pin-pitch representative value △ γ；
Step 5, the pin-pitch representative value △ γ for being obtained using step 4, in original-gray image, to each pin Around center, half through adopting intermediate value Otsu algorithm on the pin neighborhood for △ γ/2, re-starts the segmentation in chip pin region, Obtain pin local binary image；
The Otsu algorithm is maximum between-cluster variance algorithm；
Step 6, the pin local binary image to obtaining in step 5 carry out morphology opening operation and closed operation, to go Except untreated miscellaneous point and cavity in pin local binary image, pin mask images are obtained；
Step 7, the pin mask images for obtaining in step 6 and original-gray image are carried out and operation, complete chip The extraction of pin.
Beneficial effects of the present invention are：
The present invention solves existing main flow pin extracting method under the conditions of inhomogeneous illumination, and poor robustness, precision are low Problem.The uneven situation of the image brightness distribution for causing is disturbed particular for the scattering of chip surface light, proposed by the present invention dynamic The combination algorithm of state threshold value and Otsu solves pin " over-segmentation " and " less divided " problem that traditional algorithm is caused, and improves Extraction effect, robustness and precision that pin is extracted, reduce the complexity of pin extraction.
Description of the drawings
Fig. 1 is the flow chart of the chip pin extracting method in specific embodiment one under the conditions of inhomogeneous illumination；
Fig. 2 is that the industrial component image in step one to industrial camera collection carries out the original gradation figure after gray processing process Picture
Fig. 3 is in specific embodiment two, and the original-gray image under the conditions of inhomogeneous illumination is divided using dynamic threshold Cut the two-value pinoutss picture for obtaining；
Fig. 4 is in specific embodiment five, each the pin neighborhood to original-gray image under the conditions of inhomogeneous illumination, adopts The pin local binary image for obtaining after local segmentation is carried out with intermediate value Ostu algorithm.
Fig. 5 in step 7 is carried out and operation pin mask images with original-gray image, and the chip pin for obtaining is carried Take image.
Specific embodiment
Specific embodiment one：In conjunction with Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5, present embodiment, one kind of present embodiment are described Chip pin extracting method detailed process under the conditions of inhomogeneous illumination is：
Step one, the industrial component image to industrial camera collection carry out gray processing process, obtain original-gray image；
Step 2, the original-gray image obtained by step one, carry out chip pin using dynamic threshold segmentation method Thick extraction, obtains two-value pinoutss picture；
Step 3, the two-value pinoutss picture to obtaining in step 2 carry out connected component labeling, and to marked connected domain Screened, retained the connected domain of the number more than predetermined threshold value T of contained pixel in connected domain, obtained is screened through connected domain Two-value pinoutss picture afterwards；
Each in step 4, the two-value pinoutss picture after connected domain screening obtained by step 3 is effectively connected Domain, asks for its center；The spacing of each adjacent leads is calculated, using the minima in spacing as pin-pitch representative value △ γ；
Step 5, the pin-pitch representative value △ γ for being obtained using step 4, in original-gray image, to each pin Around center, half through adopting intermediate value Otsu algorithm on the pin neighborhood for △ γ/2, re-starts the segmentation in chip pin region, Obtain pin local binary image；
The Otsu algorithm is maximum between-cluster variance algorithm；
Step 6, the pin local binary image to obtaining in step 5 carry out morphology opening operation and closed operation, to go Except untreated miscellaneous point and cavity in pin local binary image, pin mask images are obtained；
Step 7, the pin mask images for obtaining in step 6 and original-gray image are carried out and operation, complete chip The extraction of pin.
Specific embodiment two：In conjunction with Fig. 3, present embodiment is described, present embodiment is different from specific embodiment one It is：The original-gray image in the step 2, step one being obtained, carries out chip pin using dynamic threshold segmentation method Slightly extract, detailed process is：
Step 2 one, mean filter is carried out to original-gray image, obtain the gray level image after mean filter, said process Mathematic(al) representation as follows:
In formula, the pixel value at position (x, y) place in the gray level image after T (x, y) expression mean filter, w (s, t) is one Size is the filter template of m × n, and f (x, y) represents the gray value at position (x, y) place in original-gray image；M, n are filtering The size of device template, value is positive integer；
In above formula, a, b meet：
A=(m-1)/2；B=(n-1)/2
Gray level image after step 2 two, the mean filter obtained by step 2 one carries out dynamic threshold segmentation, obtains two Value pinoutss picture, the mathematic(al) representation of said process is as follows：
In formula, I (x, y) represents the pixel value at position (x, y) place in two-value pinoutss picture, and C is constant, to be according to collection figure Illumination condition during picture is preassigned.
Other steps and parameter are identical with specific embodiment one.
Specific embodiment three：Present embodiment from unlike specific embodiment one or two：The constant C value is 80.
Other steps and parameter are identical with specific embodiment one or two.
Specific embodiment four：Unlike one of present embodiment and specific embodiment one to three：The step 3 In connected component labeling is carried out to the two-value pinoutss picture that obtains in step 2, and marked connected domain is screened, retains In connected domain, the number of contained pixel is more than the connected domain of predetermined threshold value T；Detailed process is：
Step 3 one, carry out connected component labeling to the two-value pinoutss picture for obtaining slightly is extracted in step 2, now on image Each pin be considered as a connected domain, and be assigned one and unique number；
Step 3 two, each connected domain is analyzed：Judge the length at number of pixels, i.e. edge for including in the connected domain Degree, if more than predetermined threshold value T；
Step 3 three, if greater than predetermined threshold value T, then retain the connected domain；Otherwise, the connected domain is rejected, I.e. on binaryzation pinoutss picture, the pixel value of all pixels that the connected domain is included is all set to 0.
One of other steps and parameter and specific embodiment one to three are identical.
Specific embodiment five：Unlike one of present embodiment and specific embodiment one to four：The step 4 In step 3 is obtained through connected domain screening after bianry image in each effective connected domain, ask for its center； The spacing of each adjacent leads is calculated, using the minima in spacing as pin-pitch representative value △ γ；Detailed process is：
Step 4 one, to each the effective connected domain in the bianry image after connected domain is screened, ask for its centre bit Put, wherein, the center position coordinates of i-th complete gray scale chip pinComputing formula is as follows：
In formula：NiFor the number of pixels of i-th complete gray scale chip pin, (xk,yk) draw for i-th complete gray scale chip The coordinate position of k-th pixel that foot includes, i and k are positive integer；
Step 4 two, each the complete gray scale chip pin center for obtaining for step 4 one, calculate each complete ash Spacing between degree chip pin center, and using the minima in spacing as pin-pitch representative value △ γ.
One of other steps and parameter and specific embodiment one to four are identical.
Specific embodiment six：In conjunction with Fig. 4, present embodiment is described, present embodiment and specific embodiment one to five it Unlike one：The pin-pitch representative value △ γ for being obtained using step 4 in the step 5 is in original-gray image, right Around each pin center, half through adopting intermediate value Otsu algorithm on the pin neighborhood for △ γ/2, re-starts chip pin region Segmentation, obtain pin local binary image；Detailed process is：
Step May Day, the segmentation threshold of each chip pin region of search is asked for, the concrete mathematic(al) representation for calculating is as follows：
In formula, L represents number of greyscale levels, ω1And ω2Represent region by prospect C after the segmentation of threshold value t respectively1With background C2In The ratio of contained number of pixels, MAD1And MAD2Then represent the average absolute variance of opposed area gradation of image intermediate value；It is specifically defined As follows：
In formula, m1(t) and m2T () represents region prospect C respectively1With background C2Gray scale intermediate value；H (i) is i-th of image The ratio of number of pixels contained by gray level；T value is 0,1,2, L, L；
Respective regions are carried out threshold value by step 5 two, the segmentation threshold of the pin region of search for being obtained using step May Day Segmentation, obtains pin local binary image.
One of other steps and parameter and specific embodiment one to five are identical.
Specific embodiment seven：Unlike one of present embodiment and specific embodiment one to six：For gray-scale maps As being 8, described step May Day, the mellow lime series L that spends was 255.
One of other steps and parameter and specific embodiment one to six are identical.

Claims (7)

1. the chip pin extracting method under the conditions of a kind of inhomogeneous illumination, it is characterised in that：Under the conditions of a kind of inhomogeneous illumination Chip pin extracting method detailed process be：
Step one, the industrial component image to industrial camera collection carry out gray processing process, obtain original-gray image；
Step 2, the original-gray image obtained by step one, carry out slightly carrying for chip pin using dynamic threshold segmentation method Take, obtain two-value pinoutss picture；
Step 3, the two-value pinoutss picture to obtaining in step 2 carry out connected component labeling, and marked connected domain are carried out Screening, retains the connected domain of the number more than predetermined threshold value T of contained pixel in connected domain, obtain after connected domain screening Two-value pinoutss picture；
Each effective connected domain in step 4, the two-value pinoutss picture after connected domain screening obtained by step 3, asks Take its center；The spacing of each adjacent leads is calculated, using the minima in spacing as pin-pitch representative value △ γ；
Step 5, the pin-pitch representative value △ γ for being obtained using step 4, in original-gray image, to each pin center Surrounding half is re-started the segmentation in chip pin region, obtains through adopting intermediate value Otsu algorithm on the pin neighborhood for △ γ/2 Pin local binary image；
The Otsu algorithm is maximum between-cluster variance algorithm；
Step 6, the pin local binary image to obtaining in step 5 carry out morphology opening operation and closed operation, are drawn with removing In foot local binary image, untreated miscellaneous point and cavity, obtain pin mask images；
Step 7, the pin mask images for obtaining in step 6 and original-gray image are carried out and operation, complete chip pin Extraction.
2. chip pin extracting method according to claim 1 under the conditions of a kind of inhomogeneous illumination, it is characterised in that：Described The original-gray image in step 2, step one being obtained, carries out the thick extraction of chip pin using dynamic threshold segmentation method, Detailed process is：
Step 2 one, mean filter is carried out to original-gray image, obtain the gray level image after mean filter, the number of said process Learn expression formula as follows:
T ( x , y ) = Σ s = - a a Σ s = - b b w ( s , t ) f ( x + s , y + t ) Σ s = - a a Σ s = - b b w ( s , t )
In formula, the pixel value at position (x, y) place in the gray level image after T (x, y) expression mean filter, w (s, t) is a size For the filter template of m × n, the gray value at position (x, y) place in f (x, y) expression original-gray image；M, n are wave filter moulds The size of plate, value is positive integer；
In above formula, a, b meet：
A=(m-1)/2；B=(n-1)/2
Gray level image after step 2 two, the mean filter obtained by step 2 one carries out dynamic threshold segmentation, obtains two-value and draws Foot image, the mathematic(al) representation of said process is as follows：
I ( x , y ) = 1 | f ( x , y ) - T ( x , y ) | &GreaterEqual; C 0 | f ( x , y ) - T ( x , y ) | < C
In formula, I (x, y) represents the pixel value at position (x, y) place in two-value pinoutss picture, and C is constant.
3. chip pin extracting method according to claim 2 under the conditions of a kind of inhomogeneous illumination, it is characterised in that：Described Constant C value is 80.
4. chip pin extracting method according to claim 3 under the conditions of a kind of inhomogeneous illumination, it is characterised in that：Described Two-value pinoutss picture in step 3 to obtaining in step 2 carries out connected component labeling, and marked connected domain is sieved Choosing, retains the connected domain of the number more than predetermined threshold value T of contained pixel in connected domain；Detailed process is：
Step 3 one, carry out connected component labeling to the two-value pinoutss picture for obtaining slightly is extracted in step 2, now every on image Individual pin is considered as a connected domain, and is assigned a unique numbering；
Step 3 two, each connected domain is analyzed：Judge the length at number of pixels, i.e. edge for including in the connected domain, Whether predetermined threshold value T is more than；
Step 3 three, if greater than predetermined threshold value T, then retain the connected domain；Otherwise, the connected domain is rejected, that is, is existed On binaryzation pinoutss picture, the pixel value of all pixels that the connected domain is included is all set to 0.
5. chip pin extracting method according to claim 4 under the conditions of a kind of inhomogeneous illumination, it is characterised in that：Described Each effective connected domain in the bianry image after connected domain screening in step 4, step 3 being obtained, asks for its center Position；The spacing of each adjacent leads is calculated, using the minima in spacing as pin-pitch representative value △ γ；Detailed process For：
Step 4 one, to each the effective connected domain in the bianry image after connected domain is screened, ask for its center, its In, the center position coordinates of i-th complete gray scale chip pinComputing formula is as follows：
( x c e n t e r i , y c e n t e r i ) = ( &Sigma; k = 1 N i x k / N i , &Sigma; k = 1 N i y k / N i )
In formula：NiFor the number of pixels of i-th complete gray scale chip pin, (xk,yk) it is i-th complete gray scale chip pin bag The coordinate position of k-th pixel for containing, i and k are positive integer；
Step 4 two, each the complete gray scale chip pin center for obtaining for step 4 one, calculate each complete gray scale core Spacing between piece pin center, and using the minima in spacing as pin-pitch representative value △ γ.
6. chip pin extracting method according to claim 5 under the conditions of a kind of inhomogeneous illumination, it is characterised in that：Described The pin-pitch representative value △ γ for being obtained using step 4 in step 5, in original-gray image, to each pin center week Half is enclosed through using intermediate value Otsu algorithm on the pin neighborhood for △ γ/2, the segmentation in chip pin region being re-started, being drawn Foot local binary image；Detailed process is：
Step May Day, the segmentation threshold of each chip pin region of search is asked for, the concrete mathematic(al) representation for calculating is as follows：
t * = arg min 0 &le; t &le; L { &omega; 1 ( t ) MAD 1 + &omega; 2 ( t ) MAD 2 }
In formula, L represents number of greyscale levels, ω1And ω2Represent region by prospect C after the segmentation of threshold value t respectively1With background C2In contained The ratio of number of pixels, MAD1And MAD2Then represent the average absolute variance of opposed area gradation of image intermediate value；Be specifically defined as Under：
&omega; 1 ( t ) = &Sigma; i = 0 t h ( i ) ,
&omega; 2 ( t ) = &Sigma; i = t + 1 L h ( i ) = 1 - &omega; 1 ( t )
MAD 1 = &Sigma; i = 0 t h ( i ) &omega; 1 ( t ) | i - m 1 ( t ) | ,
MAD 2 = &Sigma; i = 0 t h ( i ) &omega; 2 ( t ) | i - m 2 ( t ) |
In formula, m1(t) and m2T () represents region prospect C respectively1With background C2Gray scale intermediate value；H (i) is i-th gray level of image The ratio of contained number of pixels；T value is 0,1,2, L, L；
Respective regions are entered row threshold division by step 5 two, the segmentation threshold of the pin region of search for being obtained using step May Day, Obtain pin local binary image.
7. chip pin extracting method according to claim 6 under the conditions of a kind of inhomogeneous illumination, it is characterised in that：Described Step May Day, mellow lime degree series L was 255.
CN201610872992.6A 2016-09-30 2016-09-30 Chip pin extraction method under non-uniform illumination condition Pending CN106447673A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610872992.6A CN106447673A (en) 2016-09-30 2016-09-30 Chip pin extraction method under non-uniform illumination condition

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610872992.6A CN106447673A (en) 2016-09-30 2016-09-30 Chip pin extraction method under non-uniform illumination condition

Publications (1)

Publication Number Publication Date
CN106447673A true CN106447673A (en) 2017-02-22

Family

ID=58171562

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610872992.6A Pending CN106447673A (en) 2016-09-30 2016-09-30 Chip pin extraction method under non-uniform illumination condition

Country Status (1)

Country Link
CN (1) CN106447673A (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107507130A (en) * 2017-08-01 2017-12-22 江苏理工学院 A kind of quickly QFN chip pins image obtains and amplification method
CN110246139A (en) * 2019-04-26 2019-09-17 清华大学深圳研究生院 Planktonic organism in-situ image ROI rapid extracting method based on dual threshold
CN110930390A (en) * 2019-11-22 2020-03-27 郑州智利信信息技术有限公司 Chip pin missing detection method based on semi-supervised deep learning
CN112005624A (en) * 2018-05-31 2020-11-27 雅马哈发动机株式会社 Recognition parameter optimization device, component mounting system, and recognition parameter optimization method
CN112419224A (en) * 2020-07-17 2021-02-26 宁波智能装备研究院有限公司 Spherical pin chip positioning method and system
CN112419225A (en) * 2020-07-17 2021-02-26 宁波智能装备研究院有限公司 SOP type chip detection method and system based on pin segmentation
CN113284160A (en) * 2021-04-23 2021-08-20 北京天智航医疗科技股份有限公司 Method, device and equipment for identifying operation navigation mark bead body

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004046994A1 (en) * 2002-11-21 2004-06-03 Qinetiq Limited Histological assessment of nuclear pleomorphism
CN103035001A (en) * 2012-12-06 2013-04-10 中国科学院自动化研究所 Foundation automatic cloud detection method based on superpixel division
CN103761534A (en) * 2014-01-22 2014-04-30 哈尔滨工业大学 Method for detecting vision localization of QFP element

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004046994A1 (en) * 2002-11-21 2004-06-03 Qinetiq Limited Histological assessment of nuclear pleomorphism
CN103035001A (en) * 2012-12-06 2013-04-10 中国科学院自动化研究所 Foundation automatic cloud detection method based on superpixel division
CN103761534A (en) * 2014-01-22 2014-04-30 哈尔滨工业大学 Method for detecting vision localization of QFP element

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
王菡: "基于中值的Otsu算法在图像处理中的研究与应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107507130A (en) * 2017-08-01 2017-12-22 江苏理工学院 A kind of quickly QFN chip pins image obtains and amplification method
CN112005624B (en) * 2018-05-31 2022-04-26 雅马哈发动机株式会社 Recognition parameter optimization device, component mounting system, and recognition parameter optimization method
CN112005624A (en) * 2018-05-31 2020-11-27 雅马哈发动机株式会社 Recognition parameter optimization device, component mounting system, and recognition parameter optimization method
CN110246139A (en) * 2019-04-26 2019-09-17 清华大学深圳研究生院 Planktonic organism in-situ image ROI rapid extracting method based on dual threshold
CN110246139B (en) * 2019-04-26 2021-01-01 清华大学深圳研究生院 Method for rapidly extracting plankton in-situ image ROI (region of interest) based on double thresholds
CN110930390A (en) * 2019-11-22 2020-03-27 郑州智利信信息技术有限公司 Chip pin missing detection method based on semi-supervised deep learning
CN110930390B (en) * 2019-11-22 2020-09-22 深圳市海芯微迅半导体有限公司 Chip pin missing detection method based on semi-supervised deep learning
CN112419224A (en) * 2020-07-17 2021-02-26 宁波智能装备研究院有限公司 Spherical pin chip positioning method and system
CN112419224B (en) * 2020-07-17 2021-08-27 宁波智能装备研究院有限公司 Spherical pin chip positioning method and system
CN112419225B (en) * 2020-07-17 2021-10-12 宁波智能装备研究院有限公司 SOP type chip detection method and system based on pin segmentation
CN112419225A (en) * 2020-07-17 2021-02-26 宁波智能装备研究院有限公司 SOP type chip detection method and system based on pin segmentation
CN113284160A (en) * 2021-04-23 2021-08-20 北京天智航医疗科技股份有限公司 Method, device and equipment for identifying operation navigation mark bead body
CN113284160B (en) * 2021-04-23 2024-03-12 北京天智航医疗科技股份有限公司 Method, device and equipment for identifying surgical navigation mark beads

Similar Documents

Publication Publication Date Title
CN106447673A (en) Chip pin extraction method under non-uniform illumination condition
CN109961049B (en) Cigarette brand identification method under complex scene
Jidong et al. Recognition of apple fruit in natural environment
CN112837290B (en) Crack image automatic identification method based on seed filling algorithm
CN111091544B (en) Method for detecting breakage fault of side integrated framework of railway wagon bogie
CN107038416B (en) Pedestrian detection method based on binary image improved HOG characteristics
CN113409267B (en) Pavement crack detection and segmentation method based on deep learning
CN104376551A (en) Color image segmentation method integrating region growth and edge detection
CN102663757A (en) Semi-automatic image cutting method based on nuclear transfer
CN109447111B (en) Remote sensing supervision classification method based on subclass training samples
CN111754538B (en) Threshold segmentation method for USB surface defect detection
CN107154044B (en) Chinese food image segmentation method
CN102184404B (en) Method and device for acquiring palm region in palm image
CN104021566A (en) GrabCut algorithm-based automatic segmentation method of tongue diagnosis image
CN103198479A (en) SAR image segmentation method based on semantic information classification
CN104766316A (en) Novel lip segmentation algorithm for traditional Chinese medical inspection diagnosis
CN111126185B (en) Deep learning vehicle target recognition method for road gate scene
CN111539980A (en) Multi-target tracking method based on visible light
Han et al. An enhanced image binarization method incorporating with Monte-Carlo simulation
CN104281850A (en) Character area identification method and device
CN107578414B (en) Method for processing pavement crack image
CN109978916A (en) Vibe moving target detecting method based on gray level image characteristic matching
CN104102911A (en) Image processing for AOI (automated optical inspection)-based bullet appearance defect detection system
CN108009480A (en) A kind of image human body behavioral value method of feature based identification
CN111161264A (en) Method for segmenting TFT circuit image with defects

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20170222

WD01 Invention patent application deemed withdrawn after publication