基于单摄像机结构的视觉坐标测量原理研究
|
摘要
论文根据传统立体视觉传感器体积大、测量速度慢、成本高等缺点,研究了一种用于测量空间三维点坐标的单摄像机立体视觉传感器模型。在结构设计方面,变传统的双摄像机结构为单摄像机结构,加装两组对称的光学反射镜片,镜像出一对虚拟摄像机,可在同一个摄像机像面上采集到存在视差的两幅图像,从而恢复空间点的三维信息。论文从理论设计上和实践上进行了深入的研究和探讨,并完成了相应的实验验证。论文主要研究工作如下:
1.结合结构光传感器和立体视觉传感器的优点,研究了基于单摄像机虚拟立体视觉传感器的方法。该方法采用立体视差原理,由单个摄像机即可得到视场内所有物体特征点的三维信息。既提高了测量速度,同时又降低了传感器的成本。
2.分析了立体视觉传感器基线距与传感器测量精度的关系,为单摄像机立体视觉传感器的结构设计提供理论基础。
3.根据单摄像机立体视觉传感器的原理,对传感器的结构进行设计。详细分析了反射镜的摆放角度对基线距、传感器交汇角的影响。
4.根据传感器的结构设计,分析了系统中实摄像机和反射镜片的位置对虚拟摄像机位置的影响。
5.相对与传统立体视觉传感器,本结构的视场范围发生了较大的改变。基于此,本文详细分析了各参数对系统视场的影响;反之,在得到系统视场的基础上,对系统结构进行优化设计。
6.在立体视觉传感器标定原理的基础上,研究了单摄像机立体视觉传感器的标定方法。该标定过程中,只需对实摄像机的参数进行标定,简化了标定过程,降低了标定工作量。
According to the disadvantages of conventional vision sensor such as large volume、slow measurement and high cost, a kind of low cost single camera model used to measure coordinates in space is introduced. In structure design, the traditional double CCD cameras is changed to single CCD camera. By adding two groups of symmetrical reflectors, the single camera is imaged to a pair of virtual cameras. Then two pictures of the same object with parallax can be gathered from one CCD’s image surface. After that, the three-dimensional information of the point can be retrieved.
This dissertation’s main work is as follows:
1. Combining the advantages of structured light sensor and stereo vision sensor, a method based on single camera simulating stereo vision sensor has been studied. Based on stereoscopic vision, the three-dimensional information of the point can be retrieved by only one CCD’s camera. Not only has the measurement speed been advanced, but also has the sensor’s cost been decreased.
2. The relationship between stereo vision sensor’s baseline distance and stereo vision sensor’s measurement accuracy has been analyzed. This relationship provides the theoretical foundation of single camera stereo vision sensor’s structural design.
3. The sensor’s structure has been designed based on the theory of the single camera stereo vision sensor, and this dissertation has analyzed the baseline distance、the sensor’s cross angle infected by placement angles of the optical reflectors in detail.
4. Based on the structural design of the sensor, this dissertation has analyzed the position of the virtue camera influenced by the position of the real camera and the reflector.
5. Compared with the two CCD cameras system, the field-of-view of this sensor has been decreased greatly. This dissertation has analyzed the system’s field of view influenced by every parameter; whereas, optimum design of the system’s structure has based on the system’s filed of view.
6. The calibration of single camera stereo vision sensor has been researched based on the calibration of two CCD cameras system. In the process of the calibration, only one CCD camera needs to calibrate which has simplified the calibration process and reduced calibration workload.
1.结合结构光传感器和立体视觉传感器的优点,研究了基于单摄像机虚拟立体视觉传感器的方法。该方法采用立体视差原理,由单个摄像机即可得到视场内所有物体特征点的三维信息。既提高了测量速度,同时又降低了传感器的成本。
2.分析了立体视觉传感器基线距与传感器测量精度的关系,为单摄像机立体视觉传感器的结构设计提供理论基础。
3.根据单摄像机立体视觉传感器的原理,对传感器的结构进行设计。详细分析了反射镜的摆放角度对基线距、传感器交汇角的影响。
4.根据传感器的结构设计,分析了系统中实摄像机和反射镜片的位置对虚拟摄像机位置的影响。
5.相对与传统立体视觉传感器,本结构的视场范围发生了较大的改变。基于此,本文详细分析了各参数对系统视场的影响;反之,在得到系统视场的基础上,对系统结构进行优化设计。
6.在立体视觉传感器标定原理的基础上,研究了单摄像机立体视觉传感器的标定方法。该标定过程中,只需对实摄像机的参数进行标定,简化了标定过程,降低了标定工作量。
According to the disadvantages of conventional vision sensor such as large volume、slow measurement and high cost, a kind of low cost single camera model used to measure coordinates in space is introduced. In structure design, the traditional double CCD cameras is changed to single CCD camera. By adding two groups of symmetrical reflectors, the single camera is imaged to a pair of virtual cameras. Then two pictures of the same object with parallax can be gathered from one CCD’s image surface. After that, the three-dimensional information of the point can be retrieved.
This dissertation’s main work is as follows:
1. Combining the advantages of structured light sensor and stereo vision sensor, a method based on single camera simulating stereo vision sensor has been studied. Based on stereoscopic vision, the three-dimensional information of the point can be retrieved by only one CCD’s camera. Not only has the measurement speed been advanced, but also has the sensor’s cost been decreased.
2. The relationship between stereo vision sensor’s baseline distance and stereo vision sensor’s measurement accuracy has been analyzed. This relationship provides the theoretical foundation of single camera stereo vision sensor’s structural design.
3. The sensor’s structure has been designed based on the theory of the single camera stereo vision sensor, and this dissertation has analyzed the baseline distance、the sensor’s cross angle infected by placement angles of the optical reflectors in detail.
4. Based on the structural design of the sensor, this dissertation has analyzed the position of the virtue camera influenced by the position of the real camera and the reflector.
5. Compared with the two CCD cameras system, the field-of-view of this sensor has been decreased greatly. This dissertation has analyzed the system’s field of view influenced by every parameter; whereas, optimum design of the system’s structure has based on the system’s filed of view.
6. The calibration of single camera stereo vision sensor has been researched based on the calibration of two CCD cameras system. In the process of the calibration, only one CCD camera needs to calibrate which has simplified the calibration process and reduced calibration workload.
引文
【1】 K.G. Harding and K. Goodson, “Hybrid, high accuracy structured light profiler” SPIE Vol.728, Optics, Illuminmmation and Image Sensing for Machine Vision, 1986, 132~145.
【2】 N.C. Griswold and C.P. Yeh, “A new stereo vision model based upon the binocular fusion concept”, Comput. Vision, Graphics and Image Processing, Vol. 41, 1988, 153~171.
【3】 M. Ito and A. Ishi, “Three-view stereo analysis”, IEEE Trans on PAMI-8, No.4, 1986, 524~532.
【4】 E. Baumgartner and G. Saridis, “A ranging method based on triangulation for robotic control”, Proc. of IEEE Int. Conf. on Robotics and Automation, 1985, 86~91.
【5】 Zhu Ji-gui, Li Yan-jun, Ye Sheng-hua. Calibration of line structured light vision system based on camera’s projective center. Optics and Precision Engineering, 2005,13(5): 584~591.
【6】 Armin Gruen, Thomas S. Huang. Calibration and orientation of cameras in computer vision.(M) Berlin ; New York : Springer, 2001,199-205.
【7】 王东峰、张丽飞、刘小军等。基于广义特征点匹配的全自动图像配准,电子与信息学报,2005, 27(7):1013~1016.
【8】 王宝光、贺忠海 、陈林才等。结构光传感器模型及特性分析。光学学报,2002,22(4):481～484.
【9】 周富强、邾继贵、杨学友等。双目立体视觉传感器的现场标定技术。仪器仪表学报,2000,21(2):142~145.
【10】张健新。双目立体视觉技术在工业检测中的应用研究,[博士学位论文],天津:天津大学,1996.7.
【11】王昉。,[硕士学位论文],天津:天津大学,1999.7.
【12】E. Gurewitz, “More on benefit of a third eye for machine stereo perception”, Proc. Int. Conf. on Pattern Recognition, Paris, Otc. 1986, 966~968.
【13】周忠谟.GPS 卫星测量原理与应用[M].北京:测绘出版社,1997.
【14】R. N. Neison and T.Y. Young, “Determining three-dimensional objects shape and orientation from a single perspective view”, Optical Engineering, Vol. 25, No. 3, 1986,394~401.
【15】R. Horaud, “New methods for 3-D objects with single perspective views”, IEEE Trans. On PAMI-9, No.3,1987, 401~412.
【16】周富强。三维在线机器视觉检测关键技术的研究,[博士学位论文],天津:天津大学,2000.1.
【17】http://publish.it168.com/cword/718.shtml
【18】http://www.microvision.com.cn/shownews.asp?newsid=929
【19】刘常杰、邾继贵,等。汽车白车身机器视觉检测系统。汽车工程,2000,22(6):373～376.
【20】吴斌。大型物体三维形貌数字化测量关键技术研究,[博士学位论文], 天津:天津大学,2003.
【21】I. Sobel,On calibration computer controlled camera for perceiving 3D scenes,Artificial Intelligence,1974(5):185-188.
【22】Pu. P. Isaguirre,J. Summers,A new development in camera calibration: calibrating a pair of mobile cameras,Proc. Of Int. Conf. Rob. & Auto.,1985:74-79.
【23】G. Q. Wei ,S. D. Ma,Complete two-plane camera calibration and experimental comparisons,Proc. ICCV'93,1993:439-446.
【24】Y. Yakimovsky,R. Cunningham,A system for extracting three dimensional measurements from a stereo pair of TV cameras,Computer Graphics and Image Processing,1978(7):195-210
【25】H. A. Martins,J. R. Birk,R. B. Kelley,Camera models based on data from two-calibration planes,Computer Graphics and Imaging Processing,1981(17):173-180
【26】R. Y. Tsai,An efficient and accurate camera calibration technique for 3-D machine vision,Proc. CVPR'86,1986:364-374.
【27】H. Bacakoglu、M. Kamel、Senior Member,An optimized two-step camera calibration method , IEEE International Conference on Robotics and Automation,April,1997,pp.1347～1352.
【28】J. Weng,P. Cohen,M. Herniou,Calibration of stereo cameras using a non-linear distortion model,Proc. of Int. Conf. on Pattern Recognition,1990:246-253.
【29】E. L. Hall,M. B. K. Tio,C. A. Mc Pherson,et al,Curved surface measurement and recognition for robot vision , Proc. IEEE Workshop on Industrial Application of Machine Vision,1982.
【30】W. Faig,Calibration of close-range photogrammetric systems: mathematical formulation,Photogrammetric Eng. Remote sensing, 1975,41(12):1479- 1486.
【31】“Manual of Photogrammery”, Amer. Soc. Photogrammery, 1980.
【32】Decentering distortion olfenses”, Photogrammetric Eng. Remote Sensing, 1966, 444~462.
【33】郑南宁。计算机视觉与模式识别。北京:国防工业出版社,1998
【34】毛剑飞、陈利红,等。机器人视觉系统的自标定新算法。浙江大学学报,2003,37(3):283～287.
【35】Liu Xiangbin,Xiang Jianchi,“Yang Bo,A labeling algorithm based on 8-connected boundary tracking”, ComputerEngineering and Applications,2001(23): 125～12.
【36】Wang Songgui,et al,Linear Statistic Model[M]。Beijing:Higher Education Pres,1999.
【37】K. S. PARK, C. J. LIM, An efficient camera calibration method for vision-based head tracking, International Journal of Human-Computer Studies, 2000, 52(5): 879-898.
【38】Chanchal Chatterjee, Vwani P. Roychowdhury, and Edwin K. P. Chong, A Nonlinear Gauss–Seidel Algorithm for Noncoplanar and Coplanar Camera Calibration with Convergence Analysis, Computer Vision and Image Understanding , 1997, 67(1) , 58–80.
【39】Roger.Y.Tsai,A Versatile Camera Calibration Technique for High-Accuracy 3D Machine Vision Metrology Using Off-the-Shelf TV Camera and Lens. IEEE JOURNAL OF ROBOTICS AND AUTOMATION, 1987, Ra-3(4):323~344.
【40】沙吉乐。基于液晶光栅技术的三维视觉测量系统研究,[博士学位论文],天津:天津大学,2003.7.
【41】邹凤娇,苏显渝,李美菊。基于共面点的摄像机线性标定法,光电工程,2005,32(4):70~74.
【42】王浩。基于视觉的形貌测量拼接技术研究,[硕士学位论文],天津:天津大学,2005.1.
【43】S. D. Ma,G. Q. Wei,A self-calibration technique for active vision system, Proc. ACCV'95,1995:628-631.
【2】 N.C. Griswold and C.P. Yeh, “A new stereo vision model based upon the binocular fusion concept”, Comput. Vision, Graphics and Image Processing, Vol. 41, 1988, 153~171.
【3】 M. Ito and A. Ishi, “Three-view stereo analysis”, IEEE Trans on PAMI-8, No.4, 1986, 524~532.
【4】 E. Baumgartner and G. Saridis, “A ranging method based on triangulation for robotic control”, Proc. of IEEE Int. Conf. on Robotics and Automation, 1985, 86~91.
【5】 Zhu Ji-gui, Li Yan-jun, Ye Sheng-hua. Calibration of line structured light vision system based on camera’s projective center. Optics and Precision Engineering, 2005,13(5): 584~591.
【6】 Armin Gruen, Thomas S. Huang. Calibration and orientation of cameras in computer vision.(M) Berlin ; New York : Springer, 2001,199-205.
【7】 王东峰、张丽飞、刘小军等。基于广义特征点匹配的全自动图像配准,电子与信息学报,2005, 27(7):1013~1016.
【8】 王宝光、贺忠海 、陈林才等。结构光传感器模型及特性分析。光学学报,2002,22(4):481～484.
【9】 周富强、邾继贵、杨学友等。双目立体视觉传感器的现场标定技术。仪器仪表学报,2000,21(2):142~145.
【10】张健新。双目立体视觉技术在工业检测中的应用研究,[博士学位论文],天津:天津大学,1996.7.
【11】王昉。,[硕士学位论文],天津:天津大学,1999.7.
【12】E. Gurewitz, “More on benefit of a third eye for machine stereo perception”, Proc. Int. Conf. on Pattern Recognition, Paris, Otc. 1986, 966~968.
【13】周忠谟.GPS 卫星测量原理与应用[M].北京:测绘出版社,1997.
【14】R. N. Neison and T.Y. Young, “Determining three-dimensional objects shape and orientation from a single perspective view”, Optical Engineering, Vol. 25, No. 3, 1986,394~401.
【15】R. Horaud, “New methods for 3-D objects with single perspective views”, IEEE Trans. On PAMI-9, No.3,1987, 401~412.
【16】周富强。三维在线机器视觉检测关键技术的研究,[博士学位论文],天津:天津大学,2000.1.
【17】http://publish.it168.com/cword/718.shtml
【18】http://www.microvision.com.cn/shownews.asp?newsid=929
【19】刘常杰、邾继贵,等。汽车白车身机器视觉检测系统。汽车工程,2000,22(6):373～376.
【20】吴斌。大型物体三维形貌数字化测量关键技术研究,[博士学位论文], 天津:天津大学,2003.
【21】I. Sobel,On calibration computer controlled camera for perceiving 3D scenes,Artificial Intelligence,1974(5):185-188.
【22】Pu. P. Isaguirre,J. Summers,A new development in camera calibration: calibrating a pair of mobile cameras,Proc. Of Int. Conf. Rob. & Auto.,1985:74-79.
【23】G. Q. Wei ,S. D. Ma,Complete two-plane camera calibration and experimental comparisons,Proc. ICCV'93,1993:439-446.
【24】Y. Yakimovsky,R. Cunningham,A system for extracting three dimensional measurements from a stereo pair of TV cameras,Computer Graphics and Image Processing,1978(7):195-210
【25】H. A. Martins,J. R. Birk,R. B. Kelley,Camera models based on data from two-calibration planes,Computer Graphics and Imaging Processing,1981(17):173-180
【26】R. Y. Tsai,An efficient and accurate camera calibration technique for 3-D machine vision,Proc. CVPR'86,1986:364-374.
【27】H. Bacakoglu、M. Kamel、Senior Member,An optimized two-step camera calibration method , IEEE International Conference on Robotics and Automation,April,1997,pp.1347～1352.
【28】J. Weng,P. Cohen,M. Herniou,Calibration of stereo cameras using a non-linear distortion model,Proc. of Int. Conf. on Pattern Recognition,1990:246-253.
【29】E. L. Hall,M. B. K. Tio,C. A. Mc Pherson,et al,Curved surface measurement and recognition for robot vision , Proc. IEEE Workshop on Industrial Application of Machine Vision,1982.
【30】W. Faig,Calibration of close-range photogrammetric systems: mathematical formulation,Photogrammetric Eng. Remote sensing, 1975,41(12):1479- 1486.
【31】“Manual of Photogrammery”, Amer. Soc. Photogrammery, 1980.
【32】Decentering distortion olfenses”, Photogrammetric Eng. Remote Sensing, 1966, 444~462.
【33】郑南宁。计算机视觉与模式识别。北京:国防工业出版社,1998
【34】毛剑飞、陈利红,等。机器人视觉系统的自标定新算法。浙江大学学报,2003,37(3):283～287.
【35】Liu Xiangbin,Xiang Jianchi,“Yang Bo,A labeling algorithm based on 8-connected boundary tracking”, ComputerEngineering and Applications,2001(23): 125～12.
【36】Wang Songgui,et al,Linear Statistic Model[M]。Beijing:Higher Education Pres,1999.
【37】K. S. PARK, C. J. LIM, An efficient camera calibration method for vision-based head tracking, International Journal of Human-Computer Studies, 2000, 52(5): 879-898.
【38】Chanchal Chatterjee, Vwani P. Roychowdhury, and Edwin K. P. Chong, A Nonlinear Gauss–Seidel Algorithm for Noncoplanar and Coplanar Camera Calibration with Convergence Analysis, Computer Vision and Image Understanding , 1997, 67(1) , 58–80.
【39】Roger.Y.Tsai,A Versatile Camera Calibration Technique for High-Accuracy 3D Machine Vision Metrology Using Off-the-Shelf TV Camera and Lens. IEEE JOURNAL OF ROBOTICS AND AUTOMATION, 1987, Ra-3(4):323~344.
【40】沙吉乐。基于液晶光栅技术的三维视觉测量系统研究,[博士学位论文],天津:天津大学,2003.7.
【41】邹凤娇,苏显渝,李美菊。基于共面点的摄像机线性标定法,光电工程,2005,32(4):70~74.
【42】王浩。基于视觉的形貌测量拼接技术研究,[硕士学位论文],天津:天津大学,2005.1.
【43】S. D. Ma,G. Q. Wei,A self-calibration technique for active vision system, Proc. ACCV'95,1995:628-631.