CN105023265A - Checkerboard angular point automatic detection method under fish-eye lens - Google Patents

Checkerboard angular point automatic detection method under fish-eye lens Download PDF

Info

Publication number
CN105023265A
CN105023265A CN201410174721.4A CN201410174721A CN105023265A CN 105023265 A CN105023265 A CN 105023265A CN 201410174721 A CN201410174721 A CN 201410174721A CN 105023265 A CN105023265 A CN 105023265A
Authority
CN
China
Prior art keywords
angle point
angular
checkerboard
point
comers
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
CN201410174721.4A
Other languages
Chinese (zh)
Inventor
张云洲
王少楠
杨文纶
张益凯
张翰铎
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.)
Northeastern University China
Original Assignee
Northeastern University China
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 Northeastern University China filed Critical Northeastern University China
Priority to CN201410174721.4A priority Critical patent/CN105023265A/en
Publication of CN105023265A publication Critical patent/CN105023265A/en
Pending legal-status Critical Current

Links

Abstract

A checkerboard angular point automatic detection method under a fish-eye lens comprises the steps of selecting a particular region of a checkerboard and shielding useless information to reduce angular point detection time and raise detection precision; carrying out Harris angular point detection on the particular region; adding an angular point manual extraction step in an algorithm and manually adding the hidden angular point information; adding angular point combining operation according to the total number of checkerboard angular points and angular point separation situations; as for the checkerboard which is small in checkerboard rotating angle, slight in distortion and few in angular corner, sorting the angular points in a binomial fitting manner; and as for the checkerboard distorted under the fish-eye lens, calculating the sequence of the checkerboard angular points by using a modified convex hull algorithm. Simulation software is utilized to automatically detect coordinates of the checkerboard angular points, and under the condition of ensuring detection reliability of the angular points, the detection speed and precision of the angular points are greatly raised, and the complexity of manually extracting the checkerboard angular points is removed.

Description

X-comers automatic testing method under fish eye lens
Technical field
The present invention relates to field of machine vision, be specifically related to the X-comers automatic testing method under a kind of fish eye lens.
Background technology
Camera calibration, as the most important condition of vision measurement, is intended to the transformation relation set up between image coordinate system and space coordinates.Camera calibration generally comprises the detection of image characteristic point and camera parameters solves 2 steps.Cross-hatch pattern picture, as typical uncalibrated image, is widely used in camera calibration, and its Corner Detection is one of hot issue becoming computer vision also.For cross-hatch pattern picture, the task that image characteristic point detects not only will extract the image coordinate of each angle point in cross-hatch pattern picture, also will set up relation one to one between each image angle point and space angle point.
At present, X-comers detect delay mainly concentrates on gridiron pattern picture that is undistorted or that distort little, and automatic angular-point detection method can be divided into two classes: straight-line detection method and corner detection operator method.Straight-line detection method first carries out edge extracting, straight-line detection and matching, then asks the accurate coordinates of angle point by two straight-line intersections.The sequence of its unique point is simple, but when lens distortion, testing result can produce comparatively big error.Corner detection operator method utilizes the feature of angle point near zone greyscale transformation to design different corner detection operator, as SUSAN operator and Harris operator etc., response size according to corner detection operator determines angle point, but because various corner detection operator can only detect Pixel-level, need to carry out sub-pixel angle point grid further.Therefore, corner detection operator method need solve corner detection operator simultaneously, response obtains threshold value, subpixel corner detecting 3 problems.
For the Corner Detection problem of the cross-hatch pattern picture distorting comparatively serious under wide-angle lens, mostly take now the method for manual extraction angle point, namely manually mark the angle point on cross-hatch pattern picture.The method that the paper of rare automatic Corner Detection utilizes all more complicated and accuracy rate is not high, detect gridiron pattern deformation rectangle if any utilizing heuristic search and then determine the method for angle point.Other calibration tool case (as Zhang Zhengyou calibration tool case) can carry out wide-angle, flake or panoramic camera and demarcate.First user needs the gridiron pattern picture taking several different visual angles, then four of cross-hatch pattern picture angle points is manually extracted the input as camera calibration together with curved parameter.Such tool box just can automatically by other Corner Detection out, and utilize corner detection operator to carry out position optimization.But supposition is not made to the shape of camera lens in these tool boxes, and very undesirable for the fish eye lens image detect effect that distortion is serious.
In sum, traditional checkerboard angle point detection process directly cannot carry out the automatic Corner Detection of Severe distortion cross-hatch pattern picture under fish eye lens.
Summary of the invention
For the deficiency that prior art exists, the present invention propose a kind of utilize the corner detection operator of improvement and algorithm of convex hull to carry out fish eye lens under X-comers automatic testing method, the method is simple and accuracy rate is high, and the automatic Corner Detection based on the method is applicable to the cross-hatch pattern picture of different distortion degree and different angle point quantity.
X-comers automated process under fish eye lens of the present invention, comprises improvement Harris angular-point detection method and carries out the extraction of Pixel-level X-comers, introduce algorithm of convex hull and make improvements and sort to unordered angle point.Particularly: to distorting under fish eye lens, X-comers carries out Harris Corner Detection.Utilize known X-comers number herein, calculate the angle point that should merge and direct adjacent corner points to be merged.May exist reflective for picture or the extraneous factor impact such as block and cause indivedual Corner Detection not phenomenon out, the method adds masking operations removal Independent Point to be affected and adopts automatic and manual operation angle point grid to combine, and reliably solves distortion chessboard Corner Detection problem.The angle point utilizing above-mentioned improvement Harris Corner Detection to go out is unordered, introduces algorithm of convex hull herein and sorts to X-comers.Under the present invention utilizes fish eye lens there is cambered outwards phenomenon in picture, detected the angle point of outermost in X-comers by algorithm of convex hull, angle point order in recycling geometric relationship determination convex closure.For the anglec of rotation and the less chessboard angle point image that distorts, the present invention proposes mode part angle point being adopted to curve quadratic fit, is first divided into groups by angle point by ordinate large I, and recycling horizontal ordinate size realizes the sequence to overall angle point.
X-comers automatic testing method under fish eye lens, comprises the steps:
Step one: because gridiron pattern picture surrounding there will be external environment, affect tessellated detection, therefore needed to add " mask process ", masks garbage to reduce the Corner Detection time and to improve accuracy of detection before doing Harris Corner Detection.
Step 2: Harris Corner Detection is carried out to specific region.(1) correlation matrix M is calculated to each pixel; (2) the Harris angle point response of every pixel is calculated; (3) within the scope of w × w, find maximum point, if the response of Harris angle point is greater than threshold values, be then considered as angle point.
Step 3: due to uneven illumination or seriously reflective, in picture, angle point information can be hidden sometimes, for increasing reliability, need add the step of manual extraction angle point, manually being added by the angle point information be hidden.
Step 4: due to wide-angle lens limited resolution, is amplified to Pixel-level and can finds " pixel separation " phenomenon by cross-hatch pattern picture.Harris Corner Detection Algorithm asks for each pixel transverse and longitudinal Coordination difference in image, and whether utilize CRF criterion function to distinguish is angle point.Directly use Harris detection X-comers an angle point can be considered as two herein to treat, need to add according to X-comers sum and angle point separation case the operation merging angle point.
Step 5: be unordered with the distortion X-comers that draws of Harris algorithm improved, like this cannot corresponding world coordinates to demarcate.Only have and angle point is sorted, just can be called automatic detection angle point algorithm.Less for the gridiron pattern anglec of rotation, distortion is not very serious, angle point is not a lot of situation, following methods can be simply taked to carry out X-comers auto-sequencing: to utilize geometric coordinate position relationship and according to distortion degree, getting the minimum n of an ordinate angle point (makes the single angle point number of gridiron pattern be d, then n should between d and 2d), binomial curve matching is carried out to it, the curve in the middle of two row's angle points can be obtained.Make the point on curve be group 1, group interior angle dot sequency can be distinguished by horizontal ordinate size, remove the angle point of sorted group 1, cycling is carried out to remaining angle point, the sequence of X-comers can be obtained.
Step 6: use for reference convex closure theory of algorithm in computational geometry, in the ideal case (every one deck angle point all presents convex hull shape), can obtain the angle point that most edge is convex successively, be 4 convex closures herein, then the starting point found out in convex closure can obtain the angle point that sorts.To Graham scanning algorithm be utilized to solve convex closure problem referred to as algorithm of convex hull in the present invention.
Solution for first angle point problem identificatioin in algorithm of convex hull (only knows first point in convex closure, just can know the order of peripheral each point): connect adjacent corner points, slope and limit, the limit angle on limit corresponding to each angle point can be obtained; get four points that four maximum angles of limit, limit angle are corresponding; i.e. four angle points of convex closure, get from the nearest angle point of initial point as first point.After calculating the angle point of the not strict convex closure of outermost in X-comers, these points are concentrated from X-comers and removes, proceed the convex closure ordered steps improved.Just the X-comers automatically detected can be obtained thus.
beneficial effect:
The present invention utilizes simulation software automatically to detect X-comers coordinate, under the prerequisite ensureing Corner Detection reliability, greatly refer to speed and the precision of Corner Detection, has broken away from the triviality of manual extraction X-comers.
accompanying drawing illustrates:
The method flow diagram of Fig. 1 specific embodiment of the invention;
Gridiron pattern schematic diagram under the fish eye lens of Fig. 2 specific embodiment of the invention;
Fig. 3 specific embodiment of the invention Harris Corner Detection design sketch;
The Harris Corner Detection design sketch that Fig. 4 specific embodiment of the invention is improved;
Fig. 5 specific embodiment of the invention is based on the angle point sequence schematic diagram of binomial fitting;
Fig. 6 specific embodiment of the invention algorithm of convex hull detects peripheral angle point schematic diagram;
Fig. 7 specific embodiment of the invention is improved convex closure and is detected peripheral angle point schematic diagram;
Fig. 8 specific embodiment of the invention is based on the angle point sequence schematic diagram improving algorithm of convex hull.
Embodiment
Elaborate to specific embodiment of the invention below in conjunction with accompanying drawing, method flow diagram is shown in Fig. 1.
The present invention illustrates for 8 × 10 gridiron pattern fault images under fish eye lens, sees Fig. 2.
Step one: because gridiron pattern picture surrounding there will be external environment, affect tessellated detection, therefore needed to mask garbage to improve accuracy of detection before doing Harris Corner Detection.In X-comers picture, peripheral white portion is larger, selects the operation of specific region more easy to operate.Ideally, only need to choose picture periphery 4 points.
Step 2: Harris Corner Detection is carried out to specific region, sees Fig. 3.Concrete operations comprise: (1) calculates correlation matrix M to each pixel; (2) the Harris angle point response of every pixel is calculated.R is the matrix of 576 × 702 herein; (3) within the scope of w × w, find maximum point, if the response of Harris angle point is greater than threshold values, be then considered as angle point.
Threshold selection is determined according to X-comers number p herein.Due to the existence of " angle point separation " phenomenon, the angle point number of setting should be greater than actual X-comers number, p=100 herein.
The angular coordinate (coordinate order is from left to right) that Harris extracts specific region:
x 358 283 269 298 201 194 343 257 215 369
y 348 334 406 267 390 452 421 470 322 281
x 190 143 238 145 148 192 248 310 286 247
y 500 432 581 436 376 504 519 532 591 521
x 241 242 160 326 143 230 284 295 335 296
y 558 555 316 484 480 261 594 569 597 567
x 192 425 190 108 146 105 311 106 146 119
y 541 364 538 420 484 415 214 460 518 312
x 330 110 370 322 111 432 365 351 154 174
y 600 466 538 614 367 298 542 572 552 261
x 150 374 410 191 387 84 115 390 178 236
y 523 227 432 565 494 359 503 492 258 600
x 192 151 111 346 432 83 193 246 434 93
y 586 547 498 576 243 449 569 212 247 309
x 121 118 250 276 415 80 193 136 373 197
y 533 528 209 609 543 403 214 261 600 211
x 460 475 124 95 260 375 431 320 478 150
y 442 376 551 515 176 189 207 179 263 220
x 440 156 481 158 87 169 395 123 105 359
y 494 216 319 571 481 186 573 223 266 616
Step 3: due to uneven illumination or seriously reflective, in picture, angle point information can be hidden sometimes, for increasing reliability, the angle point information be hidden manually need be extracted and being joined by the angle point of extraction in the angle point matrix improving Harris extraction.
Step 4: due to wide-angle lens limited resolution, is amplified to Pixel-level and can finds " pixel separation " phenomenon by cross-hatch pattern picture.Harris Corner Detection Algorithm asks for each pixel transverse and longitudinal Coordination difference in image, and whether utilize CRF criterion function to distinguish is angle point.Directly use Harris detection X-comers an angle point can be considered as two herein to treat, need to add according to X-comers sum and angle point separation case the operation merging angle point, angle point grid effect as shown in Figure 4.
Step 1: calculate the detection angle a=100 that counts and add the difference c=24 that manual extraction angle point number b=4 and actual corners count.
Step 2: calculate each angle point spacing and ascendingly to sort.
Step 3: merged by nearest 12 pairs of angle points, getting mid point is new angle point.
The angular coordinate (coordinate order is from left to right) that the Harris improved extracts:
x 190.4631 535.8506 225.5332 178.7731 348 334 406 267 390 452
y 141.3118 101.9908 471.821 213.5775 358 283 269 298 201 194
x 421 470 322 281 502 434 581 376 520 532
y 343 257 215 369 191 144 238 148 248 310
x 593 557 316 484 482 261 568 599 540 364
y 285 242 160 326 145 230 296 333 191 425
x 418 214 463 521 312 540 614 367 298 574
y 107 311 108 148 119 368 322 111 432 349
x 550 260 227 432 567 493 359 501 600 586
y 153 176 374 410 192 389 84 113 236 192
x 245 449 211 309 531 609 543 403 213 261
y 433 83 248 93 120 276 415 80 195 136
x 600 442 376 551 515 176 189 207 179 263
y 373 460 475 124 95 260 375 431 320 478
x 218 494 319 571 481 186 573 223 266 616
y 153 440 481 158 87 169 395 123 105 359
Step 5: with less for the gridiron pattern anglec of rotation, distortion is not very serious, angle point is not a lot of situation, following methods can be simply taked to carry out X-comers auto-sequencing: to utilize geometric coordinate position relationship and according to distortion degree, 18 angle points getting ordinate minimum carry out binomial curve matching, the curve in the middle of two row's angle points can be obtained, i.e. the curve of Fig. 5 Green circle formation.The point of order on green curve is group 1, can distinguish group interior angle dot sequency, remove the angle point of sorted group 1, carry out cycling, can obtain the sequence of X-comers to remaining angle point by horizontal ordinate size.
Step 6: use for reference convex closure theory of algorithm in computational geometry, utilizes the X-comers order of function convhull calculating in convex closure in Matlab.Because Harris Corner Detection is accurate not, the interference of the unequal factor of gridiron pattern plate, cause X-comers outermost can not form strict convex, and algorithm of convex hull be using ordinate (or horizontal ordinate) minimum o'clock as first angle point, clockwise (or counterclockwise) sequence, herein different first the angle point difference that convex closure can be caused to detect of the gridiron pattern anglec of rotation.These situations all will be taken into account.
What likely lack for X-comers improves one's methods: because X-comers number is certain, the angle point number of outermost is determined thereupon, therefore the angle point number lacked after carrying out algorithm of convex hull is certain, asks for the distance of angle point to convex closure, is added by nearest several angle points.First time only detects outermost 31 angle points when running algorithm of convex hull, as shown in Figure 6.Need that 1 angle point nearest from convex closure is joined convex closure point to concentrate.Because algorithm of convex hull sorts automatically, the angle point added must be inserted into the specific position of convex closure collection, this problem can by calculate from the nearest point of angle point to be added and utilize relative position relation to judge its insertion position solves.In addition, in matlab, in convhull function, the situation of multi-point and common-line is not considered in the calculating of convex closure, needs the point of conllinear to add in convex closure herein.Herein check point concentrate whether have angle point to be formed at convex closure graph edge on, if had, add point set.
Solution for first angle point problem identificatioin in algorithm of convex hull (only knows first point in convex closure, just can know the order of peripheral each point): connect adjacent corner points, slope and limit, the limit angle on limit corresponding to each angle point can be obtained, get four points that four maximum angles of limit, limit angle are corresponding, i.e. four angle points of convex closure, the ground floor outermost obtained 4 angle points are (225.5332 471.821) (616.0000 359.0000) (535.8506 101.9908) (190.4631 141.3118), get from the nearest angle point (190.4631 141.3118) of initial point as first point, outermost Corner Detection also sorts effect as shown in Figure 7.
Outermost 32 angular coordinates (coordinate order is from left to right) after sequence:
x 190.4631 223 266 309 359 403 449 481
y 141.3118 123 105 93 84 80 83 87
x 515 535.8506 551 571 586 600 609 614
y 95 101.9908 124 158 192 236 276 322
x 616 600 573 543 494 442 376 319
y 359 373 395 415 440 460 475 481
x 263 225.5332 207 189 179 176 178.7731 186
y 478 471.821 431 375 320 260 213.5775 169
After calculating the angle point of the not strict convex closure of outermost in X-comers, concentrated from X-comers by these 32 points and remove, proceed the convex closure ordered steps improved, outermost obtains the order of 32,24,16,8 angle points successively.Just can automatically be detected thus and the X-comers sorted, angle point detects automatically and the effect that sorts is shown in Fig. 8.
X-comers coordinate (coordinate order is from left to right) after auto-sequencing:
x 190.4631 223 266 309 359 403 449 481 515 535.8506
y 141.3118 123 105 93 84 80 83 87 95 101.9908
x 551 571 586 600 609 614 616 600 573 543
y 124 158 192 236 276 322 359 373 395 415
x 494 442 376 319 263 225.5332 207 189 179 176
y 440 460 475 481 478 471.821 431 375 320 260
x 178.7731 186 218 261 312 367 418 463 501 531
y 213.5775 169 153 136 119 111 107 108 113 120
x 550 567 581 593 599 574 540 493 432 364
y 153 192 238 285 333 349 368 389 410 425
x 298 245 227 214 211 213 260 316 376 434
y 432 433 374 311 248 195 176 160 148 144
x 482 521 540 557 568 532 484 421 348 281
y 145 148 191 242 296 310 326 343 358 369
x 267 261 322 390 452 502 520 470 406 334
y 298 230 215 201 194 191 248 257 269 283
In sum, improvement Harris is utilized to detect X-comers and introduce algorithm of convex hull achieves X-comers automatic detection to the method that angle point sorts.The method, under the prerequisite ensureing Corner Detection reliability, greatly refer to speed and the precision of Corner Detection, has broken away from the triviality of manual extraction X-comers.

Claims (2)

1. the X-comers automatic testing method under fish eye lens, adopts improvement Harris algorithm to carry out the extraction of Pixel-level X-comers, introduces algorithm of convex hull and make improvements and sort to unordered angle point; Concrete steps comprise: (1) selects specific region in gridiron pattern, mask garbage and improve accuracy of detection to reduce the Corner Detection time; (2) Harris Corner Detection is carried out to specific region; (3) add the step of manual extraction angle point in algorithm, the angle point information be hidden manually is added; (4) add according to X-comers sum and angle point separation case the operation merging angle point; (5) less for the gridiron pattern anglec of rotation, distortion is comparatively light and angle point is not a lot of gridiron pattern, and the mode of binomial fitting can be utilized to sort to angle point; (6) for the gridiron pattern that distorts under most of fish eye lens, X-comers order is calculated by the algorithm of convex hull improved.
2. the X-comers automatic testing method under fish eye lens according to claim 1, is characterized in that: the described mode of binomial fitting that utilizes sorts to angle point, and the angle point sequence for qualified distortion cross-hatch pattern picture is very easy; Be incorporated into by algorithm of convex hull in large distortion scaling board angle point extraction problem automatically, and utilize geometric relationship and Relative slope and angular dimension to improve algorithm of convex hull, solving common Angular Point Extracting Method cannot to the problem of angle point sequence.
CN201410174721.4A 2014-04-29 2014-04-29 Checkerboard angular point automatic detection method under fish-eye lens Pending CN105023265A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410174721.4A CN105023265A (en) 2014-04-29 2014-04-29 Checkerboard angular point automatic detection method under fish-eye lens

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410174721.4A CN105023265A (en) 2014-04-29 2014-04-29 Checkerboard angular point automatic detection method under fish-eye lens

Publications (1)

Publication Number Publication Date
CN105023265A true CN105023265A (en) 2015-11-04

Family

ID=54413207

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410174721.4A Pending CN105023265A (en) 2014-04-29 2014-04-29 Checkerboard angular point automatic detection method under fish-eye lens

Country Status (1)

Country Link
CN (1) CN105023265A (en)

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105701776A (en) * 2016-01-07 2016-06-22 武汉精测电子技术股份有限公司 Lens distortion correcting method and system used for automatic optical detection
CN107680035A (en) * 2017-09-29 2018-02-09 广东中星微电子有限公司 A kind of parameter calibration method and device, server and readable storage medium storing program for executing
CN107767422A (en) * 2017-09-18 2018-03-06 深圳开阳电子股份有限公司 A kind of fish-eye bearing calibration, device and portable terminal
CN107909085A (en) * 2017-12-01 2018-04-13 中国科学院长春光学精密机械与物理研究所 A kind of characteristics of image Angular Point Extracting Method based on Harris operators
CN107993263A (en) * 2017-10-27 2018-05-04 深圳市易成自动驾驶技术有限公司 Viewing system automatic calibration method, automobile, caliberating device and storage medium
CN107993268A (en) * 2017-12-26 2018-05-04 广东工业大学 A kind of method and system of Camera Self-Calibration
CN108198222A (en) * 2018-01-29 2018-06-22 大连东软信息学院 A kind of wide-angle lens calibration and image correction method
CN108986721A (en) * 2018-06-14 2018-12-11 武汉精测电子集团股份有限公司 A kind of test pattern generation method for display panel detection
CN111047633A (en) * 2018-10-12 2020-04-21 富华科精密工业（深圳）有限公司 Monocular distance measuring device
CN111754461A (en) * 2020-05-28 2020-10-09 江苏理工学院 Method and device for positioning image character area of semiconductor chip
CN111798422A (en) * 2020-06-29 2020-10-20 福建汇川物联网技术科技股份有限公司 Checkerboard angular point identification method, device, equipment and storage medium
CN112923918A (en) * 2021-01-26 2021-06-08 南京理工大学 Visual pose measurement method based on improved checkerboard target

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101814186A (en) * 2010-02-04 2010-08-25 上海交通大学 Method utilizing curve-fitting to calibrate radial distortion of camera
CN103345755A (en) * 2013-07-11 2013-10-09 北京理工大学 Chessboard angular point sub-pixel extraction method based on Harris operator

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101814186A (en) * 2010-02-04 2010-08-25 上海交通大学 Method utilizing curve-fitting to calibrate radial distortion of camera
CN103345755A (en) * 2013-07-11 2013-10-09 北京理工大学 Chessboard angular point sub-pixel extraction method based on Harris operator

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
中国指挥控制学会火力与指挥控制专业委员会: "《神经中枢 2013 网络时代的火力与指挥控制》", 31 October 2013, 北京:国防工业出版社 *
冯焕飞: "一种高精度特征点检测和相机标定方法", 《中小企业管理与科技(上旬刊)》 *
张裕 等: "基于Harris算法的黑白棋盘格角点检测", 《计算机应用与软件》 *
杨文纶 等: "基于逆向多项式模型的鱼眼图像畸变矫正研究", 《第六届全国大学生创新创业年会》 *
熊会元 等: "基于凸包的棋盘格角点自动识别与定位方法", 《中山大学学报（自然科学版）》 *

Cited By (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105701776A (en) * 2016-01-07 2016-06-22 武汉精测电子技术股份有限公司 Lens distortion correcting method and system used for automatic optical detection
CN107767422A (en) * 2017-09-18 2018-03-06 深圳开阳电子股份有限公司 A kind of fish-eye bearing calibration, device and portable terminal
CN107767422B (en) * 2017-09-18 2022-04-05 深圳开阳电子股份有限公司 Fisheye lens correction method and device and portable terminal
CN107680035A (en) * 2017-09-29 2018-02-09 广东中星微电子有限公司 A kind of parameter calibration method and device, server and readable storage medium storing program for executing
CN107680035B (en) * 2017-09-29 2021-08-17 广东中星微电子有限公司 Parameter calibration method and device, server and readable storage medium
CN107993263B (en) * 2017-10-27 2021-07-06 深圳市易成自动驾驶技术有限公司 Automatic calibration method for panoramic system, automobile, calibration device and storage medium
CN107993263A (en) * 2017-10-27 2018-05-04 深圳市易成自动驾驶技术有限公司 Viewing system automatic calibration method, automobile, caliberating device and storage medium
CN107909085A (en) * 2017-12-01 2018-04-13 中国科学院长春光学精密机械与物理研究所 A kind of characteristics of image Angular Point Extracting Method based on Harris operators
CN107993268A (en) * 2017-12-26 2018-05-04 广东工业大学 A kind of method and system of Camera Self-Calibration
CN107993268B (en) * 2017-12-26 2021-11-30 广东工业大学 Camera self-calibration method and system
CN108198222B (en) * 2018-01-29 2021-09-03 大连东软信息学院 Wide-angle lens calibration and image correction method
CN108198222A (en) * 2018-01-29 2018-06-22 大连东软信息学院 A kind of wide-angle lens calibration and image correction method
CN108986721B (en) * 2018-06-14 2021-06-04 武汉精测电子集团股份有限公司 Detection graph generation method for display panel detection
CN108986721A (en) * 2018-06-14 2018-12-11 武汉精测电子集团股份有限公司 A kind of test pattern generation method for display panel detection
CN111047633A (en) * 2018-10-12 2020-04-21 富华科精密工业（深圳）有限公司 Monocular distance measuring device
CN111047633B (en) * 2018-10-12 2023-08-18 深圳富联富桂精密工业有限公司 Monocular distance measuring device
CN111754461A (en) * 2020-05-28 2020-10-09 江苏理工学院 Method and device for positioning image character area of semiconductor chip
CN111754461B (en) * 2020-05-28 2024-03-01 江苏理工学院 Method and device for positioning image character area of semiconductor chip
CN111798422A (en) * 2020-06-29 2020-10-20 福建汇川物联网技术科技股份有限公司 Checkerboard angular point identification method, device, equipment and storage medium
CN111798422B (en) * 2020-06-29 2023-10-20 福建汇川物联网技术科技股份有限公司 Checkerboard corner recognition method, device, equipment and storage medium
CN112923918A (en) * 2021-01-26 2021-06-08 南京理工大学 Visual pose measurement method based on improved checkerboard target

Similar Documents

Publication Publication Date Title
CN105023265A (en) Checkerboard angular point automatic detection method under fish-eye lens
CN109271937B (en) Sports ground marker identification method and system based on image processing
JP5699788B2 (en) Screen area detection method and system
CN110097542B (en) Method and device for detecting chip bubbles and storage medium
EP2811423A1 (en) Method and apparatus for detecting target
CN107248159A (en) A kind of metal works defect inspection method based on binocular vision
CN104517276A (en) Checker corner detection method
CN102209974A (en) Feature value extracting device, object identification device, and feature value extracting method
CN103727930A (en) Edge-matching-based relative pose calibration method of laser range finder and camera
CN109727239A (en) Based on SURF feature to the method for registering of inspection figure and reference map
CN106327464A (en) Edge detection method
CN105718931B (en) System and method for determining clutter in acquired images
CN105096299A (en) Polygon detection method and device
CN102236894A (en) Circle detection method and device
CN103745221A (en) Two-dimensional code image correction method
CN105787912B (en) Classification-based step type edge sub-pixel positioning method
CN109815822A (en) Inspection figure components target identification method based on Generalized Hough Transform
US10074551B2 (en) Position detection apparatus, position detection method, information processing program, and storage medium
CN109493384A (en) Camera position and orientation estimation method, system, equipment and storage medium
CN104574312A (en) Method and device of calculating center of circle for target image
CN108090492B (en) Contour detection method based on scale clue suppression
CN105678737A (en) Digital image corner point detection method based on Radon transform
Zhang et al. A new algorithm for accurate and automatic chessboard corner detection
CN103679713A (en) Two-dimensional image registration method for partially matched images
CN114119644A (en) Template matching method based on edge features

Legal Events

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

Application publication date: 20151104