精密角度基准下的多相机定位系统高精度标定
|
摘要
在视觉定位测量领域中的大尺寸测量、运动追踪、三维重建、视觉定位中,针对多相机定位系统中相机之间无公共视场或公共视场较小时系统标定困难、方法繁琐、精度低等问题,提出了一种基于精密二轴转台的多相机定位系统一体化标定方法。利用二轴转台提供角度基准,当转台一次转过所有相机视场时,各个相机依次拍摄标定图片,求解出各个相机内参以及各相机到转台的外参,利用转台坐标系中转计算相机之间的外参。整个标定过程由程序控制,实现了多相机系统采图标定的集成化、自动化,降低了标定工作量。分析了多相机定位系统的标定原理,并进行了实验验证。两相机内参重投影误差在0.17 pixel以内,系统定位精度在1 mm以内。结果表明,所提方法切实可行,精度较高,可操作性强,可应用于各种无公共视场或公共视场较小的多相机定位系统标定过程。
As for the large size measurement, motion tracking, 3 D reconstruction and visual positioning in the field of visual positioning measurement, there exist many problems of system calibration difficulty, method complexity,low precision and so on for a multi-camera positioning system with no public or less public field of view. Thus an integrated calibration method for a multi-camera positioning system is proposed based on a precision two-axis turntable, in which the two-axis turntable provides an angle reference. When the turntable passes through the fields of view of all cameras at once, each camera shoots sequentially the calibration images. The internal parameters of each camera and the external parameters from each camera to the turntable are solved. The turntable coordinates are used to calculate the external parameters of cameras. The whole calibration process is controlled by programming.The integration and automation of figure-collection calibration for the multi-camera system are achieved, which greatly reduces the workload of calibration. The calibration principle of this multi-camera positioning system is analyzed and verified by experiments as well. The reprojection errors of the internal parameters of two cameras are less than 0.17 pixel and the system positioning accuracy is less than 1 mm. The results show that the proposed method is feasible, accurate and operable, which can be applied in the calibration process of a multi-camera positioning system with no public or small public field of view.
As for the large size measurement, motion tracking, 3 D reconstruction and visual positioning in the field of visual positioning measurement, there exist many problems of system calibration difficulty, method complexity,low precision and so on for a multi-camera positioning system with no public or less public field of view. Thus an integrated calibration method for a multi-camera positioning system is proposed based on a precision two-axis turntable, in which the two-axis turntable provides an angle reference. When the turntable passes through the fields of view of all cameras at once, each camera shoots sequentially the calibration images. The internal parameters of each camera and the external parameters from each camera to the turntable are solved. The turntable coordinates are used to calculate the external parameters of cameras. The whole calibration process is controlled by programming.The integration and automation of figure-collection calibration for the multi-camera system are achieved, which greatly reduces the workload of calibration. The calibration principle of this multi-camera positioning system is analyzed and verified by experiments as well. The reprojection errors of the internal parameters of two cameras are less than 0.17 pixel and the system positioning accuracy is less than 1 mm. The results show that the proposed method is feasible, accurate and operable, which can be applied in the calibration process of a multi-camera positioning system with no public or small public field of view.
引文
[1] Zhao C Y. Research on multi-cameras calibration method based on spherical targets[D]. Hefei:Hefei University of Technology, 2016.赵成运.基于立体靶标的多摄像机标定方法研究[D].合肥:合肥工业大学,2016.
[2] Yang F, Liu W, Zhang Y, et al. Binocular camera calibration method combined with the four collinear constraints[J]. Acta Optica Sinica, 2016, 36(7):0715001.杨帆,刘巍,张洋,等.结合四角共线约束的大视场双目相机标定方法[J].光学学报,2016, 36(7):0715001.
[3] Li H H, Wang J D. A research on camera calibration technique[J]. Optical Instruments, 2007, 29(4):7-12.李洪海,王敬东.摄像机标定技术研究[J].光学仪器,2007, 29(4):7-12.
[4] Yang B W, Zhang L Y, Ye N. Camera calibration technique of wide-area vision measurement[J]. Acta Optica Sinica, 2012, 32(9):0915001.杨博文,张丽艳,叶南,等.面向大视场视觉测量的摄像机标定技术[J].光学学报,2012, 32(9):0915001.
[5] Li Q Q. The Research on calibration of the nonmetric digital camera[D]. Chengdu:SouthwestJiaotong University, 2016.李倩倩.非量测数码相机标定技术的分析与研究[D].成都:西南交通大学,2016.
[6] Tsai R Y. An efficient and accurate camera calibration technique for 3D machine vision[J].Proceedings of Computer Vision&Pattern Recognition, 1986:364-374.
[7] Zhang Z Y. A flexible new technique for camera calibration[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(11):1330-1334.
[8] Ni A W. Study on three dimension reconstruction technique based on stereo vision[D]. Nanjing:Nanjing University of Aeronautics and Astronautics,2009.倪爱伟.基于双目立体视觉的三维重建技术研究[D].南京:南京航空航天大学,2009.
[9] Wang G, Shang Y, Guan B L, et al. Flexible calibration of setting relation of a multi-camera rig for non-overlapping views[J]. Chinese Journal of Lasers,2017, 44(6):0604004.王刚,尚洋,关棒磊,等.无重叠视场多相机组安装关系的灵活标定[J].中国激光,2017, 44(6):0604004.
[10] Hu X K, Bi Y W. Stereo vision calibration method of no common view cameras[J]. Journal of Yantai University, 2010, 23(2):138-141.胡潇琨,毕远伟.一种无公共视场双摄像机位置关系的求取方法[J].烟台大学学报(自然科学与工程版),2010, 23(2):138-141.
[11] Lu Y N, Wan Z J, Wang X J, et al. Solution to relative position of cameras without public FOV[J].Journal of Applied Optics, 2017, 38(3):400-405.鲁亚楠,万子敬,王向军.一种无公共视场相机位置关系的求解方法[J].应用光学,2017, 38(3):400-405.
[12] Yang Z X. Multi-camera calibration technique research and its application[D]. Changsha:Hunan University, 2011.杨振先.多摄像机标定技术研究及其应用[D].长沙:湖南大学,2011.
[13] Gu G H, Wang J J, Chen Q, et al. Camera parameter calibration based on two-dimensional rotating platform[J]. Optics&Precision Engineering, 2017, 25(7):1890-1899.顾国华,王佳节,陈钱,等.二维旋转平台下的相机参数标定[J].光学精密工程,2017, 25(7):1890-1899.
[14] Jiang G W, Chao Z C, Fu S H,et al. High-accurate camera calibration technique based on controllable rotation[J]. Acta Optica Sinica, 2010, 30(5):1308-1314.姜广文,晁志超,伏思华,等.基于可控旋转的像机高精度标定技术[J].光学学报,2010, 30(5):1308-1314.
[15] Gao X, Zhang T, Liu Y, et al. Visual SLAM fourteen lectures:From theory to practice[M].Beijing:Electronic Industry Press, 2017.高翔,张涛,刘毅,等.视觉SLAM十四讲:从理论到实践[M].北京:电子工业出版社,2017.
[16] Zhu J G, Yu Z J. The principle of vision metrology[M]. Beijing:China Machine Press, 2012.邾继贵,于之靖.视觉测量原理与方法[M].北京:机械工业出版社,2012.
[17] Zhang M, Jin L X, Li G N, et al. Camera distortion calibration method based on straight line characteristics[J]. Acta Optica Sinica, 2015, 35(6):0615001.张敏,金龙旭,李国宁,等.基于直线特征的摄像机镜头畸变标定方法[J].光学学报,2015, 35(6):0615001.
[18] Zhu J, Li X F, Xu Y X. Camera calibration technique based on active vision[J]. Acta Optica Sinica, 2010,30(5):1297-1303.朱嘉,李醒飞,徐颖欣.摄像机的一种主动视觉标定方法[J].光学学报,2010, 30(5):1297-1303.
[19] More J J. The Levenberg-Marquardt algorithm:Implementation and theory[J]. Lecture Notesin Mathematics, 1977, 630:105-116.
[20] Zhao Y, Huang M, Zhang Q. V-STARS based photogrametry system and its application in paraboloid antenna testing[J]. Journal of Hebei University of Technology, 2012, 41(3):6-9.赵艳,黄敏,张强.V-STARS摄影测量系统及其在抛物面天线测量中的应用[J].河北工业大学学报,2012, 41(3):6-9.
[21] LI D M. The principle and application of V-STARS photographic surveying system[J].Water Conservancy&Electric Power Machinery, 2006, 28(10):26-27.李东明.V-STARS摄影测量系统的原理与应用[J].水利电力机械,2006, 28(10):26-27.
[22] Huang G P, Qin G Q, Lu C J. Study on the key technologies of digital close range industrial photogrammetry and applications[J]. Journal of Astronautic Metrology and Measurement, 2009,29(2):5-9.黄桂平,钦桂勤,卢成静.数字近景摄影大尺寸三坐标测量系统V-STARS的测试与应用[J].宇航计测技术,2009, 29(2):5-9.
[23] Zhang X, Cheng D Q, Li W. High precision calibration of vision measurement system in largeFOV based on virtual 3D target[J]. Optics&Precision Engineering, 2017, 25(4):891-899.张曦,程东勤,李伟.构建虚拟立体靶标的大视场高精度视觉标定[J].光学精密工程,2017, 25(4):891-899.
[24] Liu J W, Liang J, Liang X H, et al. Industrial vision measuring system for large dimension work-pieces[J]. Optics Precision Engineering, 2010, 18(1):126-134.刘建伟,梁晋,梁新合,等.大尺寸工业视觉测量系统[J].光学精密工程,2010, 18(1):126-134.
[25] Horaud R, Conio B, Leboulleux O, et al. An analytic solution for the perspective 4-point problem[J]. Computer Vision, Graphics, and Image Processing, 1989, 47(1):33-44.
[26] Fischler M A, Bolles R C. Random sample consensus:A paradigm for model fitting with applications to image analysis and automated cartography[J]. Readings in Computer Vision, 1987:726-740.
[2] Yang F, Liu W, Zhang Y, et al. Binocular camera calibration method combined with the four collinear constraints[J]. Acta Optica Sinica, 2016, 36(7):0715001.杨帆,刘巍,张洋,等.结合四角共线约束的大视场双目相机标定方法[J].光学学报,2016, 36(7):0715001.
[3] Li H H, Wang J D. A research on camera calibration technique[J]. Optical Instruments, 2007, 29(4):7-12.李洪海,王敬东.摄像机标定技术研究[J].光学仪器,2007, 29(4):7-12.
[4] Yang B W, Zhang L Y, Ye N. Camera calibration technique of wide-area vision measurement[J]. Acta Optica Sinica, 2012, 32(9):0915001.杨博文,张丽艳,叶南,等.面向大视场视觉测量的摄像机标定技术[J].光学学报,2012, 32(9):0915001.
[5] Li Q Q. The Research on calibration of the nonmetric digital camera[D]. Chengdu:SouthwestJiaotong University, 2016.李倩倩.非量测数码相机标定技术的分析与研究[D].成都:西南交通大学,2016.
[6] Tsai R Y. An efficient and accurate camera calibration technique for 3D machine vision[J].Proceedings of Computer Vision&Pattern Recognition, 1986:364-374.
[7] Zhang Z Y. A flexible new technique for camera calibration[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(11):1330-1334.
[8] Ni A W. Study on three dimension reconstruction technique based on stereo vision[D]. Nanjing:Nanjing University of Aeronautics and Astronautics,2009.倪爱伟.基于双目立体视觉的三维重建技术研究[D].南京:南京航空航天大学,2009.
[9] Wang G, Shang Y, Guan B L, et al. Flexible calibration of setting relation of a multi-camera rig for non-overlapping views[J]. Chinese Journal of Lasers,2017, 44(6):0604004.王刚,尚洋,关棒磊,等.无重叠视场多相机组安装关系的灵活标定[J].中国激光,2017, 44(6):0604004.
[10] Hu X K, Bi Y W. Stereo vision calibration method of no common view cameras[J]. Journal of Yantai University, 2010, 23(2):138-141.胡潇琨,毕远伟.一种无公共视场双摄像机位置关系的求取方法[J].烟台大学学报(自然科学与工程版),2010, 23(2):138-141.
[11] Lu Y N, Wan Z J, Wang X J, et al. Solution to relative position of cameras without public FOV[J].Journal of Applied Optics, 2017, 38(3):400-405.鲁亚楠,万子敬,王向军.一种无公共视场相机位置关系的求解方法[J].应用光学,2017, 38(3):400-405.
[12] Yang Z X. Multi-camera calibration technique research and its application[D]. Changsha:Hunan University, 2011.杨振先.多摄像机标定技术研究及其应用[D].长沙:湖南大学,2011.
[13] Gu G H, Wang J J, Chen Q, et al. Camera parameter calibration based on two-dimensional rotating platform[J]. Optics&Precision Engineering, 2017, 25(7):1890-1899.顾国华,王佳节,陈钱,等.二维旋转平台下的相机参数标定[J].光学精密工程,2017, 25(7):1890-1899.
[14] Jiang G W, Chao Z C, Fu S H,et al. High-accurate camera calibration technique based on controllable rotation[J]. Acta Optica Sinica, 2010, 30(5):1308-1314.姜广文,晁志超,伏思华,等.基于可控旋转的像机高精度标定技术[J].光学学报,2010, 30(5):1308-1314.
[15] Gao X, Zhang T, Liu Y, et al. Visual SLAM fourteen lectures:From theory to practice[M].Beijing:Electronic Industry Press, 2017.高翔,张涛,刘毅,等.视觉SLAM十四讲:从理论到实践[M].北京:电子工业出版社,2017.
[16] Zhu J G, Yu Z J. The principle of vision metrology[M]. Beijing:China Machine Press, 2012.邾继贵,于之靖.视觉测量原理与方法[M].北京:机械工业出版社,2012.
[17] Zhang M, Jin L X, Li G N, et al. Camera distortion calibration method based on straight line characteristics[J]. Acta Optica Sinica, 2015, 35(6):0615001.张敏,金龙旭,李国宁,等.基于直线特征的摄像机镜头畸变标定方法[J].光学学报,2015, 35(6):0615001.
[18] Zhu J, Li X F, Xu Y X. Camera calibration technique based on active vision[J]. Acta Optica Sinica, 2010,30(5):1297-1303.朱嘉,李醒飞,徐颖欣.摄像机的一种主动视觉标定方法[J].光学学报,2010, 30(5):1297-1303.
[19] More J J. The Levenberg-Marquardt algorithm:Implementation and theory[J]. Lecture Notesin Mathematics, 1977, 630:105-116.
[20] Zhao Y, Huang M, Zhang Q. V-STARS based photogrametry system and its application in paraboloid antenna testing[J]. Journal of Hebei University of Technology, 2012, 41(3):6-9.赵艳,黄敏,张强.V-STARS摄影测量系统及其在抛物面天线测量中的应用[J].河北工业大学学报,2012, 41(3):6-9.
[21] LI D M. The principle and application of V-STARS photographic surveying system[J].Water Conservancy&Electric Power Machinery, 2006, 28(10):26-27.李东明.V-STARS摄影测量系统的原理与应用[J].水利电力机械,2006, 28(10):26-27.
[22] Huang G P, Qin G Q, Lu C J. Study on the key technologies of digital close range industrial photogrammetry and applications[J]. Journal of Astronautic Metrology and Measurement, 2009,29(2):5-9.黄桂平,钦桂勤,卢成静.数字近景摄影大尺寸三坐标测量系统V-STARS的测试与应用[J].宇航计测技术,2009, 29(2):5-9.
[23] Zhang X, Cheng D Q, Li W. High precision calibration of vision measurement system in largeFOV based on virtual 3D target[J]. Optics&Precision Engineering, 2017, 25(4):891-899.张曦,程东勤,李伟.构建虚拟立体靶标的大视场高精度视觉标定[J].光学精密工程,2017, 25(4):891-899.
[24] Liu J W, Liang J, Liang X H, et al. Industrial vision measuring system for large dimension work-pieces[J]. Optics Precision Engineering, 2010, 18(1):126-134.刘建伟,梁晋,梁新合,等.大尺寸工业视觉测量系统[J].光学精密工程,2010, 18(1):126-134.
[25] Horaud R, Conio B, Leboulleux O, et al. An analytic solution for the perspective 4-point problem[J]. Computer Vision, Graphics, and Image Processing, 1989, 47(1):33-44.
[26] Fischler M A, Bolles R C. Random sample consensus:A paradigm for model fitting with applications to image analysis and automated cartography[J]. Readings in Computer Vision, 1987:726-740.