摘要
自动光学检测技术具有检测方法先进、检测速度快、自动化程度高等优点,可在线完成物体表面空间三维坐标的测量。本文采用线结构光视觉技术建立了微小三维尺寸(0.01-10mm)自动光学检测系统。
论文研究的主要内容和创新技术有:
1.通过引入光平面和测量坐标系的夹角参数,建立了传感器的DLT改进数学模型和多项式模型。
2.研究了传感器两步标定方法,第一步将数学模型中的参数标定完成,第二步确定光平面与测量坐标系的夹角的数值。该方法可直接构建传感器的测量坐标系,过程简单,标定结果可靠,适合现场实施。
3.对传感器标定和安装过程中可能产生的误差进行了分析。提出了对标定构建的测量坐标系与实际测量坐标系偏斜造成的误差进行修正的模型,研究了使用一标准块对修正模型参数进行校准的方法。
4.研究了光强自适应调整方法,应用于研制了电路板锡膏三维测量系统,保证了光条图像的效果,降低图像处理复杂度的同时提高了光条中心提取的准确度。对该系统进行了重复性测量实验,结果表明系统的高度测量精度可达±1?m,并在现场进行了实际测量实验。
5.研究了椭圆弧光条图像的径向灰度重心提取的方法,应用于研制了微小直径高精度测量系统,提高了系统的测量精度。对该系统进行了重复性测量实验,结果表明系统的测量精度可达±1.4?m。使用该系统对发动机旋转叶片边缘曲率半径进行了实际测量。
Automatic Optical Inspection (AOI) has advantages of advanced inspectionmethod, high measuring speed and high degree of automation. It can be applied toobtain 3D coordinate of the object surface points at the scene. Based on structuredlight vision technology, an AOI system for 3D micro size (0.01-1mm) measurement isproposed in this dissertation.
The major contribution and innovations of this dissertation are as follows:
1. Two mathematical models of the sensor are studied. One is the improvedmodel based on Direct Linear Transform (DLT). The other is the polynomial model.And the angle between the light plane and the measuring coordinate system isintroduced.
2. A two-step calibration method for the mathematical models is studied. In thefirst step, the parameters of the mathematical model are confirmed. In the second step,the angle between the light plane and the measuring coordinate system is calculated.This method can build the measuring coordinate system directly. The process of themethod is simple and stable. It is suitable for on-site calibration.
3. The errors, which maybe occurred during the calibration or the installationprocess of the sensor, are analyzed. The correction model is proposed, which canreduce the error caused by the deflection between the measuring coordinateconstructed by the calibration and the actual one. The calculation method using astandard part for the parameters in the correction model is studied.
4. The adaptive adjustment method of laser light intensity is proposed, which isapplied in the developed high-precision measuring system for 3D size of solder pastesprinted on the PCBs. By using the method, the light-stripe image effect is guaranteed.And the accuracy of the light-stripe center extraction is improved while thecomplexity of image processing is reduced. The repeatability precision of the systemwas tested. The experimental result proved that the measuring precision along theheight direction is±1?m. The system was also tested on-site.
5. A novel grayscale barycenter extraction algorithm along radial directions isproposed, which is applied in the developed high-precision measuring system formicro diameter. The accuracy of the measurement result can be improved obviouslyby using the algorithm. The repeatability precision of the system was tested. The experimental result proved that the measuring precision is±1.4?m. At the scene, thesystem was applied to measure the curvature radius of metal blades.
引文
[1]廖强,周忆,米林,等.机器视觉在精密测量中的应用.重庆大学学报(自然科学版), 2002, 25(6): 1~4.
[2]崔秋生.多层互连基板的自动光学检测技术.混合微电子技术, 1991, 2(4):19~22.
[3]邹十践.以信息化带动我国装备制造业的发展.建筑机械化, 2002, (1):24~25.
[4]练元坚.发展装备制造业的分类思考.机电产品开发与创新,2001, (70):4~9.
[5]卢荣胜,自动光学检测技术的发展现状.红外与激光工程, 2008,37(Supplement): 120~123.
[6]西安交通大学机械工程学院. http://www.xjtuom.com/main/home/main.php.
[7]沈鑫,韩韬.自动光学检测在系泊链测量中的应用.测控技术, 2008, 27(7):91~93.
[8]刘永新,印刷电路板自动光学检测系统的设计与研究: [硕士学位论文],北京;北京交通大学, 2007. 12.
[9]罗兵,章云. SMT焊膏印刷质量自动光学检测.电子质量, 2005, (12): 30~32.
[10]鲜飞. AOI技术在SMT生产上的应用.电子测试, 2008, (160): 86~90.
[11]吴福培,张宪民,邝泳聪,等. SMT炉后晶体管类元件无铅焊点的自动光学检测算法.焊接学报, 2009, 30(7): 5~8.
[12]季秀橄,章云,曾散璐. SMT焊膏印刷质量AOI技术的研究.自动化技术与应用, 2007, 26(4): 114~116.
[13]鲜飞. 3D AOI应对焊膏印刷的新挑战.印制电路信息, 2007, (10): 67~69.
[14]罗兵,曾歆懿,季秀霞. SMT产品缺陷及相应的AOI检测方法.电子质量,2006, (6): 35~37.
[15]刘家豪.计算机视觉在微构件三维检测中的应用研究: [硕士学位论文],浙江;浙江大学, 2003. 1.
[16] D.A.Forsyth and J.Ponce. Computer Vision: A Modern Approach, Prentice Hall,2002.
[17] R.Klette, K.Schluns and A.Koschan. Computer Vision Three-Dimensional Datafrom Images, Springer-Verlag Singapore Pte.Ltd., 1998.
[18] Zeiss homepage, http://www.zeiss.com.cn/.
[19]张凤林.空间自由曲面测量技术的动态与发展(一).现代计量测试, 1998,(5): 3~6.
[20] Sheppard C J R, Matthews H J. The extended-focus, auto-focus and surfaceprofiling techniques of confocal microscopy. Mod Opt, 1986, 35(1): 145~154.
[21] Minsky M. Microscopy apparatus. U S Patent: 3013467, 1961-12-19.
[22] Zinser G. Topographic measurements at the fundus with the heidelberg retinatomography.视觉の科学, 1992, 13(4): 1~9.
[23] Wilson T. Depth response of scanning microscope. Optic, 1989, 81(3):113~118.
[24] Sheppard C J R,Gu M. Approximation to the three-dimensional optical transferfunction. J. Opt. Soc. Am. (A), 1991, 8(4): 692~694.
[25]田维坚,陈波.并行共焦三维检测方法的理论分析.光学学报, 1999,19(10): 1381~1385.
[26] Xiao G Q, Kino G S. A real-time confocal scanning optical microscope. SPIE,1987, 809: 107~113.
[27] Andres K. Visualization in biomedical microscopies-3D imaging and computerapplications. New York: VCH Publishers, 1992. 109~142.
[28] Yin Shizhuo, Li Jiang, Song Minho. Surface profile measurement using aunique microcube-based system. Optics Communications, 1999, 168: 1~6.
[29]顾敏,王桂英.共焦显微术的三维成像原理.北京:新时代出版社, 2000.
[30]刘文文,余晓芬,余卿,等.用于微结构探测的并行共焦光学结构的研究.仪器仪表学报, 2009, 30(6): 1266~1270.
[31]孔兵,王昭,谭玉山.利用共焦成像原理实现微米级的三维轮廓测量.西安交通大学学报, 2001, 35(11): 1151~1154.
[32]高兴宇,萧泽新,伍世荣.线结构光共焦显微成像技术的实验研究.光学技术, 2006, 32(Supplement): 173~175.
[33] Adachi M. Phase-shift algorithm for white-light interferometry insensitive to alinear error in phase shift increment, Proceedings of SPIE, 2005, 604806: 1~9.
[34] Sinclair M B, Boer M P, Corwin A D. Long-working-distance incoherent-lightinterference microscope. Applied Optics, 2005, 44(36): 7714~772.
[35]郭彤,胡春光,陈津平,等. MEMS显微干涉测量系统中相移器性能的研究.传感技术学报. 2006, 19(5): 1500~1508.
[36] Tung C H, Huang C H, Kao C F, et al. Fast surface profiling usingmonochromatic phase and fringe order in white-light interferometry.Proceedings of SPIE, the International Society for Optical Engineering, 2006,6292: 1F.1~1F.9.
[37] Sugiyama M, Ogawa H, Kitagawa K, et al. Single-shot surface profiling by localmodel fitting, Applied Optics , 2006, 45(31): 7991~8005.
[38]胡晓东,齐浩,刘钢,等.基于傅里叶干涉法的微结构离面运动测量,光电子·激光, 2007, 18(8): 966~969.
[39]徐永祥.显微干涉术在测量微小表面几何参数中的关键技术研究: [博士学位论文],南京;南京理工大学, 2007.9.
[40]戴蓉.基于垂直扫描工作台的白光干涉表面形貌测量系统研究: [博士学位论文],武汉;华中科技大学, 2007.9.
[41]胡春光,胡晓东,徐临燕,等.复杂微结构三维形貌测量方法的研究.光电子·激光, 2006, 17(1): 98~101.
[42]罗钧,黄俊.基于机器视觉的二维小尺寸精密测量系统.计算机测量与控制,2007, 15(1): 11-13.
[43]贺秋伟.基于计算机视觉的微小尺寸精密检测理论与技术研究: [博士学位论文],吉林;吉林大学, 2007.6.
[44]伍济钢.薄片零件尺寸机器视觉检测系统关键技术研究: [博士学位论文],武汉;华中科技大学, 2008.5.
[45] Schreier HW, Garcia D, Sutton MA. Advances in light microscope stereo vision.Experimental Mechanics, 2004, 44(3): 278~288.
[46]王跃宗,刘京会,李德胜,等.一种新的SLM显微立体视觉畸变矫正模型.光电工程, 2005, 32(6): 72~75.
[47]刘胜.旋转体视显微检测理论及装置研究: [博士学位论文],浙江;浙江大学, 2003.12.
[48]张效栋.便携式数字相位光栅投影全貌测量系统的研究: [博士学位论文],天津;天津大学, 2006.12.
[49]尹庆吉,唐立新.彩色光栅投影3D模型恢复中的条纹图像处理.计算机与数学工程, 2007, 35(11): 85~88.
[50]潘伟,赵毅,阮雪榆.相移法在光栅投影测量中的应用.应用光学, 2003,24(4): 45~49.
[51]梁晋,肖振中,臧顺来,等.外差式多频相移技术的三维光学点云测量研究.锻压技术, 2008, 33(1): 143~147.
[52]章毓晋,图像处理和分析(上,下),清华大学出版社,2001。
[53]李晓莹,马炳和,苑伟政.基于光切法的微结构三维尺寸测量.航空精密制造技术, 1998, 34(5): 38~40.
[54] Sun Changku, You Qiang. Online machine vision method for measuring thediameter and straightness of seamless steel pipes, Optical Engineering, 2001,40(11): 2565~2571.
[55]邱宇,SMT中BGA共面性在线测试方法的研究: [硕士学位论文],天津;天津大学,1999.7。
[56]王志雄,汽车车轮定位参数视觉检测方法的研究: [硕士学位论文],天津;天津大学,2003.1。
[57]罗维克,非接触式激光测量系统原理及应用,人造通讯,2004,11(7):26~28。
[58]杨再花等,一种基于光学三角法的形貌测量系统,光学技术,2005,31(4):662~626。
[59]胡曙光,陈静,华艳秋,等.基于激光三角法的传感器设计,2006,25(6),7~9
[60]王鹏,线结构光三维自动扫描系统关键技术的研究: [博士学位论文],天津;天津大学,2008.6。
[61]韩建栋,吕乃光,董明利,等.线结构光传感系统的快速标定方法.光学精密工程,2009,17(5):958~963。
[62]郁道银,谈恒英.工程光学.北京:机械工业出版社, 1994.
[63] R.Y.Tsai, An efficient and accurate camera calibration technique for 3Dmachine vision, In Proc. IEEE Conf. On Computer Vision and PatternRecognition, Miami Beach, FL, 1986: 364~374.
[64] Liu Bin, Sun Changku, Zhang Xiaodong, et al. 3D Laser MeasurementTechnique of Solder Paste. Proceedings of SPIE, the International Society forOptical Engineering. Vol.6623, 2008: 0Z-1 ~ 0Z-7
[65]陶立,彩色三维激光扫描测量系统关键技术的研究: [博士学位论文],天津;天津大学,2006.1。
[66]孙长库,何明霞,王鹏.激光测量技术.天津:天津大学出版社, 2008.
[67] Z.Y.Zhang, Flexible Camera Calibration by Viewing a Plane from UnknownOrientations, In Proc.ICCV99, 1999: 666~673.
[68]魏振忠,张广军,徐园.一种线结构光视觉传感器标定方法.机械工程学报,2005,41(2):210~214.
[69]徐光祜,刘立峰,曾建超,等.一种新的基于结构光的三维视觉系统标定方法.计算机学报,1995,18(6):450~456.
[70] Huynh D.Q. Calibration a structured light stripe system : a novel approach.International Journal of Computer Vision, 1999, 33(1): 73~86.
[71]邾继贵,李艳军,叶声华.基于共面标定参照物的线结构光传感器快速标定方法.中国机械工程,2006,17(2):183~186。
[72]周富强,陈酉玫,张广军.双传感器激光视觉测量系统的校准.光电子·激光,2007,18(7):828~831。
[73] Abdel-Aziz, Y.I.Karara, H.M. Direct linear transformation into object spacecoordinates in close-range photogrammetry. Proc. Symposium on Close-Rangephotogrammetry, Urbana, Illinois, p.1~18
[74] Janne Heikkila, Olli Silven, A four-step camera calibration procedure withimplicit image correction, Proceedings of the 1997 Conference on ComputerVision and Pattern Recognition (CVPR '97)
[75] Faugeras, O. D. & Toscani, G. (1987) Camera calibration for 3D computervision. Proc. International Workshop on Industrial Applications of MachineVision and Machine Intelligence, Silken, Japan, p. 240~247.
[76] Weng, J., Cohen, P. & Herniou, M. (1992) Camera calibration with distortionmodels and accuracy evaluation. IEEE Transactions on Pattern Analysis andMachine Intelligence PAMI-14(10): 965~980.
[77] Chang Ming, Lin Kao-Hui, Non-contact scanning measurement utilizing a spacemapping method, Optics and Lasers in Engineering 1999, 30: 503~12.
[78]段发阶,刘凤梅,叶声华.一种新型线结构光传感器结构参数标定方法.仪器仪表学报, 2000, 22(2): 108~110.
[79] Sun CK, Liu B, Wang P. Design and calibration method for microvisionthree-dimensional laser scanner. Journal of Vacuum Science & Technology B:Microelectronics and Nanometer Structures. 2009, 27(3): 1315~1318.
[80]陶立,彩色三维激光扫描测量成像系统的研究: [硕士学位论文],天津;天津大学,2004.1.
[81]邹永星,周仁魁,罗秀娟.一种提取图像目标边缘的新方法.光电工程,2005, 32(6): 76~78.
[82]陶立,彩色三维激光扫描测量系统关键技术的研究: [博士学位论文],天津;天津大学,2006.1。
[83] Andreas Daffertshofer, Claudine J. C. Lamotha, Onno G. Meijer, et al. PCA instudying coordination and variability: a tutorial, Clinical Biomechanics, 2004,19(4): 415~428.
[84]解则晓,张成国,张国雄,等.线结构光测头的误差补偿.仪器仪表学报,2005, 26(7): 667~725.
[85] http://local.wasp.uwa.edu.au/~pbourke/dataformats/asc/.
[86]王鹏.线结构光三维自动扫描系统关键技术的研究: [博士学位论文],天津;天津大学,2008.6。
[87]马骏.岩石表面线结构光扫描三维测量方法的研究: [硕士学位论文],天津;天津大学,2007.6。
[88]张维光,于洵,路绍军.基于傅里叶变换的干涉条纹对比度增强算法.信息技术, 2008, (10) : 24~30.
[89]徐科华,马晶,谭立英.深空光通信中图像信标的捕获和跟踪.光学学报,2006, 26(2): 176~180.
[90]王忆锋.红外图像灰度直方图统计分析的研究.红外, 2009, 30(4): 14~16.
[91]吴成玉,邰晓英.基于灰度变化统计的图像检索.计算机工程与应用, 2007,43(16): 83~86.
[92]冈萨雷斯著,阮秋琦译.数字图像处理(第二版).电子工业出版社, 2003.
[93]郝麟,冯少彤,聂守平.利用深度和灰度图像实现三维目标的识别与分类.光电子·激光, 2010, 21(2): 312~317.
[94]胡宁,孟利民.离散傅里叶变换在指纹图像分割中的应用.计算机工程与应用, 2009, 45(26): 185~186.
[95]刘斌.锡膏三维测量系统的研究: [硕士学位论文],天津;天津大学,2007.6。
[96]李渊,王淡,祁燕.基于DXF文件的面向三维重建的三视图数据结构.沈阳理工大学学报, 2006, 25(6): 17~19.
[97]于起峰,陆宏伟,刘肖琳.基于图像的精密测量与运动测量[M].北京:科学出版社, 2002.
[98]胡坤,周富强,张广军.一种快速结构光条纹中心亚像素精度提取方法.仪器仪表学报, 2006, 27(10): 1326~1329.
[99] Steger C.An unbiased detector of curvilinear structures.IEEE Transactions onPattern Analysis and machine Intelligence,1998,20(2):113~125.
[100]Andrew W. Fitzgibbon, Maurizio Pilu, Robert B. Fisher. Direct least-squaresfitting of ellipses. IEEE Transactions on Pattern Analysis and MachineIntelligence, 1999, 21(5): 476-480.
[101]彭凯,刘书桂,张雪飞,等.基于椭圆线轮廓度误差评定的椭圆提取方法.光电工程, 2006, 33(12): 101~104.
[102]W. Gander, G.H. Golub, and R. Strebel. Least-Square Fitting of Circles andEllipses. BIT, 1994, 43: 558~578.
[103]刘珂,周富强,张广军.半径约束最小二乘圆拟合方法及其误差分析.光电子·激光, 2006, 17(5): 604~607.
