基于数字相移条纹投影技术的三维测量
|
摘要
随着科学技术和工业的发展，三维测量技术在自动化生产、质量控制、机器人视觉、反求工程、CAD／CAM以及生物医学工程等方面的应用日益重要。传统的接触式测量技术存在测量力、测量时间长、需进行测头半径的补偿、不能测量弹性或脆性材料等局限性，因而不能满足现代工业发展的需要。光学非接触式三维测量技术克服了上述缺陷，成为近年来三维测量技术研究的热点。
本文描述了一种新型的非接触式的光学投影三维测量系统。该系统采用了数字相移条纹投影技术，测量具有非接触、全场、精度较高及快速等优点。在传统相移系统中，条纹图通常都是通过物理光栅得到，并且条纹图的相位移动也是由光栅的移动来实现。而本系统采用数字条纹投影系统，用计算机产生高亮度与高对比度的标准正弦投影条纹，从而提高系统的分辨率与测量精度，并用软件控制精确地实现相位移动，避免了传统相移系统带来的各种误差。
文中阐述了基于数字相移条纹投影技术的三维测量方法的基本原理。针对传统结构系统标定法中存在的问题，对传统结构进行了改进，标定中去掉了传统结构的平行度的约束要求，有较高的可操作性，提高了系统的测量精度。对摄像机镜头的非线性畸变的机理进行了研究，提出了一种提高条纹投影测量精度的摄像机镜头畸变系数的优化方法，从对参考平面的测量出发，考察其上的若干测量点，找到其变形与摄像机镜头畸变系数的函数关系，进而确定摄像机镜头畸变系数，并给出了由于镜头畸变给测量带来的误差的补偿方法。
With the development of science and technology and industry, 3D measurement techniques are increasingly important in production for automation, quality control machine vision, reversal engineering, CAD/CAM and biomedical engineering, etc. Traditional contact measurement techniques have lots of flaws, such as measurement force companying with measurement process, being time-consuming, needing compensation for radius of the test head, and being unable to measure elastic or fragile material, and so on. Therefore, it can't meet demands of the development of modern industry. Optical non-contact 3D measurement techniques have become the focus of research of 3D measurement techniques in recent years because of overcoming the defects above-mentioned.
A kind of new-type non-contact 3D measurement system is described in this paper. The system adopts digital phase-shifting and fringe projection techniques and its measurement has advantages of non-contact, the whole field, higher precision, and quite high speed. In traditional phase-shifting systems, the fringe patterns are usually abtained through the physical grating, and it is realized by the movement of the grating that phase-shifting of the fringe patterns is moved. However, the system described in this paper adopts the digital fringe projection system to produce the standard sinusoidal projection fringe with the high ratio of light and contrast by a computer, and thus raises the ratio of resolution of the system and precision of the measurement. Phase-shifting is controlled by software accurately and all kinds of errors caused by traditional phase-shifting systems will be avoided.
The basic principle of 3D measurement method based on digital phase-shifting and fringe projection techniques is illustrated in this paper. To the questions that exist in the system with traditional structure, the traditional structure is improved to remove parallel degree required in traditional structure for higher maneuverability in system calibration and higher precision of measurement. The mechanism of the non-linear distortion of the camera lens is carried on research. A kind of optimal method of the distortion coefficients of camera lens for improving the precision of the fringe projection measurement is proposed. Based on measuring the reference plane, the approach analyses the errors of some measure points on it, finds the relationship between the errors and the distortion coefficients of camera lens. Then the distortion coefficients of camera lens are determined by iterative optimization. This method
solves the problem of the error caused by the distortion of camera lens effectively and the method to compensate for the error caused by the distortion of camera lens is given.
本文描述了一种新型的非接触式的光学投影三维测量系统。该系统采用了数字相移条纹投影技术，测量具有非接触、全场、精度较高及快速等优点。在传统相移系统中，条纹图通常都是通过物理光栅得到，并且条纹图的相位移动也是由光栅的移动来实现。而本系统采用数字条纹投影系统，用计算机产生高亮度与高对比度的标准正弦投影条纹，从而提高系统的分辨率与测量精度，并用软件控制精确地实现相位移动，避免了传统相移系统带来的各种误差。
文中阐述了基于数字相移条纹投影技术的三维测量方法的基本原理。针对传统结构系统标定法中存在的问题，对传统结构进行了改进，标定中去掉了传统结构的平行度的约束要求，有较高的可操作性，提高了系统的测量精度。对摄像机镜头的非线性畸变的机理进行了研究，提出了一种提高条纹投影测量精度的摄像机镜头畸变系数的优化方法，从对参考平面的测量出发，考察其上的若干测量点，找到其变形与摄像机镜头畸变系数的函数关系，进而确定摄像机镜头畸变系数，并给出了由于镜头畸变给测量带来的误差的补偿方法。
With the development of science and technology and industry, 3D measurement techniques are increasingly important in production for automation, quality control machine vision, reversal engineering, CAD/CAM and biomedical engineering, etc. Traditional contact measurement techniques have lots of flaws, such as measurement force companying with measurement process, being time-consuming, needing compensation for radius of the test head, and being unable to measure elastic or fragile material, and so on. Therefore, it can't meet demands of the development of modern industry. Optical non-contact 3D measurement techniques have become the focus of research of 3D measurement techniques in recent years because of overcoming the defects above-mentioned.
A kind of new-type non-contact 3D measurement system is described in this paper. The system adopts digital phase-shifting and fringe projection techniques and its measurement has advantages of non-contact, the whole field, higher precision, and quite high speed. In traditional phase-shifting systems, the fringe patterns are usually abtained through the physical grating, and it is realized by the movement of the grating that phase-shifting of the fringe patterns is moved. However, the system described in this paper adopts the digital fringe projection system to produce the standard sinusoidal projection fringe with the high ratio of light and contrast by a computer, and thus raises the ratio of resolution of the system and precision of the measurement. Phase-shifting is controlled by software accurately and all kinds of errors caused by traditional phase-shifting systems will be avoided.
The basic principle of 3D measurement method based on digital phase-shifting and fringe projection techniques is illustrated in this paper. To the questions that exist in the system with traditional structure, the traditional structure is improved to remove parallel degree required in traditional structure for higher maneuverability in system calibration and higher precision of measurement. The mechanism of the non-linear distortion of the camera lens is carried on research. A kind of optimal method of the distortion coefficients of camera lens for improving the precision of the fringe projection measurement is proposed. Based on measuring the reference plane, the approach analyses the errors of some measure points on it, finds the relationship between the errors and the distortion coefficients of camera lens. Then the distortion coefficients of camera lens are determined by iterative optimization. This method
solves the problem of the error caused by the distortion of camera lens effectively and the method to compensate for the error caused by the distortion of camera lens is given.
引文
[1] F. Chen, G. M. Brown, M. Song, et al. Overview of three-dimensional shape measurement using optical methods. Optical Engineering, 2000, 39(1):10-22
[2] 苏显渝，李继陶．三维面形测量技术的新进展．物理，1996，25(10)：614-620
[3] T. C. Strand. Optical three-dimentional sensing for machine vision. Opt. Eng,1985,24(1):33-38
[4] 孙东，胡杰．三坐标测量机的最新动向与新技术．青岛大学学报，1997，12(1)：58-64
[5] 杜颖，李真，张国雄．三维曲面的光学非接触测量技术．光学精密工程，1999，7(3)：1-6
[6] 李永祺．光电技术与机器视觉在钢铁工业中的新应用．钢铁钒钛，1994，15(3)：57-62
[7] 吴春才，苏显渝．动态过程的三维面形测量．光电子·激光，1996，7(5)：273-278
[8] 林华，王晓秋．PSD及其在非接触测量中的应用．激光技术，1998，22(1)：52-54
[9] 左爱秋，阮晓东，骆涵秀等．立体视觉测量系统的三维空间重构．机电工程，1999，16(3)：6-9
[10] Jing Xu, Haiqing Zhou, Ping Zhu, etal. Stereo vision system for shape inspection of technical object videometrics. SPIE, 1992,(1820): 196-200
[11] 周富强，邻继贵，杨学友等．双目视觉传感器的现场标定技术．仪器仪表学报，2000，21(2)：142-145
[12] 李奇，冯华君，徐之海．计算机立体视觉技术综述．光学技术，1999，(5)：71-73
[13] 赵建才，闵新力，万德安．汽车形貌三维曲面测量装置的研制．机械加工与自动化，2001，(12)：18-19
[14] 高勃，王忠义，施长溪等．牙冠外形三维光学测量系统标定实验(一)——误差消除和标定曲线获得．实用口腔医学杂志，1999，15(4)：299
[15] 金观昌，高柠．自动三维度量学及其在人体头面部测量中的应用．人类学学报，1990，5：164-167
[16] 苏显渝，张冠申，陈泽先等．鞋楦三维面形光电自动测量系统．光学工程，1989(6)1：5-8
[17] 来新民，黄田，曾子平等．自由曲面逆向工程系统的研究．中国机械工
程，2000，11(7)：777-778
[18] R. A. Jarvis. Laser time-of-flight range scanner for robotic vision, IEEE Trans on PANG, 1983, 5(5):505-512
[19] 唐朝伟，梁锡昌．复杂曲面轮廓激光扫描三维视觉传感系统．光电工程，1993，20(2)：41-49
[20] 宋万忠，苏显渝．基于结构光投影的三维物体识别．光电子·激光，1999，10(1)：59-62
[21] 周利民，胡德洲，卢秉恒．激光扫描三角法测量精度因素的分析与研究．计量学报，1998，19(2)：130-135
[22] 张勇斌，卢荣胜，刘志健等．线结构光视觉测量系统的标定方法．传感器世界，2003，(8)：10-13
[23] 虞启琏，张天虎．利用结构光进行三维测量的新方法．应用光学，2000，21(1)：42-45
[24] Peisen S. Huang, F. P. Chiang. Recent advances in fringe projection technique for 3-D shape measurement. SPIE, 1999,(3783):132-142
[25] 段振广，李志惠．漫射表面的自动莫尔轮廓术．激光与光学，1990，2(6)：32-36
[26] K. W. Wong, Ying Ke, L. Michael, et al. Three-dimensional gaging with stereo computer vision. Industry Vision Metrology. SPIE, 1991, 1526:17-25
[27] 谭吉春，陈第安．基于电子光栅拍频原理的莫尔等高线测量术．光电工程，1991，18(1)：31-35
[28] 向淑兰，苏显渝，李继陶．相移扫描莫尔的付里叶描述．四川大学学报(自然科学版)，1999，36(3)：501-507
[29] 王昭，赵宏，谭玉山．频移阴影莫尔法．光学学报，1999，19(6)：816-820
[30] 曹向群，曹根宏，朱宝文．莫尔轮廓术的光学特性．光电工程，1991，18(4)：29-31
[31] M. Takeda, K. Mutoh. Fourier transform profilometry for the automatic measurement of 3-D object shapes. Applied Optics, 1983,(22):3977-3981
[32] X. Y. Su, W. S. Zhou, G. vonBally. Automated phase-measuring profiliometry using defocused projection of the Ronchi grating. Opt. Comm, 1992, 94(6):561
[33] K. Creath, J. Schmit. N-point spatial phase-measurement techniques for non-destructive testing. Opt. Lasers Eng.,1996,(24):365-379
[34] 余慧，谢铁邦，刘晓军．基于光聚焦探测法的表面粗糙度非接触测量系统．测试与设备，1998，9(3)：33-34
[35] 谢峰，谢铁邦，张新宝．表面形貌焦点跟踪测量方法的研究．计量学报，2000，21(3)：173-175
[36] 李礼夫，钟先信，陈愚．三维曲面轮廓的非接触式现代测量．实用测试技术，1996，(1)：22-26
[37] 孙桂林．全息光栅在光学测量中的应用．光学技术，1994，(1)：11-14
[38] 徐光佑，朱志刚．三维视觉传感器技术．机器人情报，1992，(6)：1-10
[39] Wansong Li, Xianyu Su, Likun Su. A practical coordinate mapping method for phase-measuring profilometry. Proc. SPIE, 1998,(3558): 125-130
[40] Shouhong Tang, Yah Huang. Fast profilometer for the automatic measurement of 3-D object shapes. Applied Optics, 1990,(29):3012～3018
[41] 郝煜栋，赵洋，李达成．光学投影式三维轮廓测量技术综述．光学技术，1998，(5)：57-63
[42] 赵宏，陈文艺，谭玉山．一种新的相位测量轮廓术．光学学报，1995，15(7)：898-901
[43] 胡逸群．用数字滤波消除散斑条纹图噪声．光学学报，1989，9(1)：61-67
[44] 范志刚，李润顺，左保军等．低对比度非均匀干涉条纹图预处理研究．工程技术，2002，28(1)：41-43
[45] 李仁举，钟约先，由志福等．三维测量系统中摄像机定标技术．清华大学学报(自然科学版)，2002，42(4)：481-483
[46] 于瀛洁，李国培，陈明仪．干涉图处理中的相位去包裹技术．宇航计测技术，2002，22(4)：49-54
[47] J. A. Quiroga, A. C. Gonzalez, E. Bernabeu. Phase-unwrapping algorithm based on an adaptive criterion. Applied Optics, 1995,34(14): 2560-2563
[48] H. Zhao, W. Chen, Y. Tan. Phase unwrapping algorithm for the measurement of three dimensional object shapes. Applied Optics, 1994,33(20):4497-4500
[49] 范志刚，李顺润．干涉条纹图的处理方法研究．光学技术，2000，26(3)：258-262
[53] 吴禄慎，任丹，吴魁．一种新的区域增长相位去包裹算法．机械工程学报，2002，38(12)：126-130
[51] K. Itoh. Analysis of phase-unwrapping algorithm. Applied Optics,1982,(21): 2470-2476
[52] R. Y. Tsai. A Veratile Camera Calibration Techique for High-Accuracy
3D Machine Vision Metrology Using Offthe Shelf TV Cameras and Lenses. IEEE Journal of Robotics and Automation, 1987, RA-3(4):323-324
[53] J. Y. Weng, P. Cohen, M. Herniou. Camera Calibraion with Distoration Models and Accuracy Evaluation. IEEE Trans on PAMI, 1992,14(10):965-980
[2] 苏显渝，李继陶．三维面形测量技术的新进展．物理，1996，25(10)：614-620
[3] T. C. Strand. Optical three-dimentional sensing for machine vision. Opt. Eng,1985,24(1):33-38
[4] 孙东，胡杰．三坐标测量机的最新动向与新技术．青岛大学学报，1997，12(1)：58-64
[5] 杜颖，李真，张国雄．三维曲面的光学非接触测量技术．光学精密工程，1999，7(3)：1-6
[6] 李永祺．光电技术与机器视觉在钢铁工业中的新应用．钢铁钒钛，1994，15(3)：57-62
[7] 吴春才，苏显渝．动态过程的三维面形测量．光电子·激光，1996，7(5)：273-278
[8] 林华，王晓秋．PSD及其在非接触测量中的应用．激光技术，1998，22(1)：52-54
[9] 左爱秋，阮晓东，骆涵秀等．立体视觉测量系统的三维空间重构．机电工程，1999，16(3)：6-9
[10] Jing Xu, Haiqing Zhou, Ping Zhu, etal. Stereo vision system for shape inspection of technical object videometrics. SPIE, 1992,(1820): 196-200
[11] 周富强，邻继贵，杨学友等．双目视觉传感器的现场标定技术．仪器仪表学报，2000，21(2)：142-145
[12] 李奇，冯华君，徐之海．计算机立体视觉技术综述．光学技术，1999，(5)：71-73
[13] 赵建才，闵新力，万德安．汽车形貌三维曲面测量装置的研制．机械加工与自动化，2001，(12)：18-19
[14] 高勃，王忠义，施长溪等．牙冠外形三维光学测量系统标定实验(一)——误差消除和标定曲线获得．实用口腔医学杂志，1999，15(4)：299
[15] 金观昌，高柠．自动三维度量学及其在人体头面部测量中的应用．人类学学报，1990，5：164-167
[16] 苏显渝，张冠申，陈泽先等．鞋楦三维面形光电自动测量系统．光学工程，1989(6)1：5-8
[17] 来新民，黄田，曾子平等．自由曲面逆向工程系统的研究．中国机械工
程，2000，11(7)：777-778
[18] R. A. Jarvis. Laser time-of-flight range scanner for robotic vision, IEEE Trans on PANG, 1983, 5(5):505-512
[19] 唐朝伟，梁锡昌．复杂曲面轮廓激光扫描三维视觉传感系统．光电工程，1993，20(2)：41-49
[20] 宋万忠，苏显渝．基于结构光投影的三维物体识别．光电子·激光，1999，10(1)：59-62
[21] 周利民，胡德洲，卢秉恒．激光扫描三角法测量精度因素的分析与研究．计量学报，1998，19(2)：130-135
[22] 张勇斌，卢荣胜，刘志健等．线结构光视觉测量系统的标定方法．传感器世界，2003，(8)：10-13
[23] 虞启琏，张天虎．利用结构光进行三维测量的新方法．应用光学，2000，21(1)：42-45
[24] Peisen S. Huang, F. P. Chiang. Recent advances in fringe projection technique for 3-D shape measurement. SPIE, 1999,(3783):132-142
[25] 段振广，李志惠．漫射表面的自动莫尔轮廓术．激光与光学，1990，2(6)：32-36
[26] K. W. Wong, Ying Ke, L. Michael, et al. Three-dimensional gaging with stereo computer vision. Industry Vision Metrology. SPIE, 1991, 1526:17-25
[27] 谭吉春，陈第安．基于电子光栅拍频原理的莫尔等高线测量术．光电工程，1991，18(1)：31-35
[28] 向淑兰，苏显渝，李继陶．相移扫描莫尔的付里叶描述．四川大学学报(自然科学版)，1999，36(3)：501-507
[29] 王昭，赵宏，谭玉山．频移阴影莫尔法．光学学报，1999，19(6)：816-820
[30] 曹向群，曹根宏，朱宝文．莫尔轮廓术的光学特性．光电工程，1991，18(4)：29-31
[31] M. Takeda, K. Mutoh. Fourier transform profilometry for the automatic measurement of 3-D object shapes. Applied Optics, 1983,(22):3977-3981
[32] X. Y. Su, W. S. Zhou, G. vonBally. Automated phase-measuring profiliometry using defocused projection of the Ronchi grating. Opt. Comm, 1992, 94(6):561
[33] K. Creath, J. Schmit. N-point spatial phase-measurement techniques for non-destructive testing. Opt. Lasers Eng.,1996,(24):365-379
[34] 余慧，谢铁邦，刘晓军．基于光聚焦探测法的表面粗糙度非接触测量系统．测试与设备，1998，9(3)：33-34
[35] 谢峰，谢铁邦，张新宝．表面形貌焦点跟踪测量方法的研究．计量学报，2000，21(3)：173-175
[36] 李礼夫，钟先信，陈愚．三维曲面轮廓的非接触式现代测量．实用测试技术，1996，(1)：22-26
[37] 孙桂林．全息光栅在光学测量中的应用．光学技术，1994，(1)：11-14
[38] 徐光佑，朱志刚．三维视觉传感器技术．机器人情报，1992，(6)：1-10
[39] Wansong Li, Xianyu Su, Likun Su. A practical coordinate mapping method for phase-measuring profilometry. Proc. SPIE, 1998,(3558): 125-130
[40] Shouhong Tang, Yah Huang. Fast profilometer for the automatic measurement of 3-D object shapes. Applied Optics, 1990,(29):3012～3018
[41] 郝煜栋，赵洋，李达成．光学投影式三维轮廓测量技术综述．光学技术，1998，(5)：57-63
[42] 赵宏，陈文艺，谭玉山．一种新的相位测量轮廓术．光学学报，1995，15(7)：898-901
[43] 胡逸群．用数字滤波消除散斑条纹图噪声．光学学报，1989，9(1)：61-67
[44] 范志刚，李润顺，左保军等．低对比度非均匀干涉条纹图预处理研究．工程技术，2002，28(1)：41-43
[45] 李仁举，钟约先，由志福等．三维测量系统中摄像机定标技术．清华大学学报(自然科学版)，2002，42(4)：481-483
[46] 于瀛洁，李国培，陈明仪．干涉图处理中的相位去包裹技术．宇航计测技术，2002，22(4)：49-54
[47] J. A. Quiroga, A. C. Gonzalez, E. Bernabeu. Phase-unwrapping algorithm based on an adaptive criterion. Applied Optics, 1995,34(14): 2560-2563
[48] H. Zhao, W. Chen, Y. Tan. Phase unwrapping algorithm for the measurement of three dimensional object shapes. Applied Optics, 1994,33(20):4497-4500
[49] 范志刚，李顺润．干涉条纹图的处理方法研究．光学技术，2000，26(3)：258-262
[53] 吴禄慎，任丹，吴魁．一种新的区域增长相位去包裹算法．机械工程学报，2002，38(12)：126-130
[51] K. Itoh. Analysis of phase-unwrapping algorithm. Applied Optics,1982,(21): 2470-2476
[52] R. Y. Tsai. A Veratile Camera Calibration Techique for High-Accuracy
3D Machine Vision Metrology Using Offthe Shelf TV Cameras and Lenses. IEEE Journal of Robotics and Automation, 1987, RA-3(4):323-324
[53] J. Y. Weng, P. Cohen, M. Herniou. Camera Calibraion with Distoration Models and Accuracy Evaluation. IEEE Trans on PAMI, 1992,14(10):965-980