多特征点拓扑确定位姿测量算法研究
|
摘要
为解决单目视觉位姿测量时,由目标特征点较多导致图像点与目标点拓扑关系未知的问题,提出了一种多特征点拓扑确定位姿测量算法。较多特征点可在目标进行大角度运动时保证足够的特征点进行位姿解算,与较少特征点相比提高测量精度。该算法将拓扑确定的过程和位姿求解的迭代过程进行嵌套,同时进行拓扑确定和位姿计算。位姿计算的迭代过程基于平行透视投影模型,不需要目标重心投影点坐标作为迭代初始值。拓扑确定的过程转化为分配问题的求解过程。每次位姿迭代的过程中进行一次拓扑确定,拓扑确定的结果可以计算更优的位姿估计。通过多位姿测量实验和精度对比实验结果证明:该算法适合大范围、高精度的位姿测量,在-120°~120°范围内,位姿测量均方根误差为0.272°。
In monocular vision pose estimation, the topological relationship between objective feature points and image feature points where there are multiple feature points is difficult to determine. An algorithm based on multiple feature points topological determination was proposed to solve this problem where the correspondences are unknown. By mounting multiple feature points on the object, enough proper feature points for pose computing were guaranteed when the object is large-scale moving,which can improve the precision of pose estimation. The algorithm nested the iteration process of topological determination and the iteration process of pose computing into one iteration loop, solving them simultaneously. The pose estimation iteration process was based on para-perspective projection model,where the coordination of the projection of the object gravity center used as the initial parameter of iteration is not needed. The iteration process of topological determination was transformed into a solution of assignment problem. Each topological determination can obtain a better pose estimation in every pose estimation iteration loop. The results of multiple poses experiment and precision comparison experiment prove that the algorithm is qualified for the high precision pose estimation of 3D object with large scale motion, with the root mean square error 0.272° in the range of-120°-120°.
In monocular vision pose estimation, the topological relationship between objective feature points and image feature points where there are multiple feature points is difficult to determine. An algorithm based on multiple feature points topological determination was proposed to solve this problem where the correspondences are unknown. By mounting multiple feature points on the object, enough proper feature points for pose computing were guaranteed when the object is large-scale moving,which can improve the precision of pose estimation. The algorithm nested the iteration process of topological determination and the iteration process of pose computing into one iteration loop, solving them simultaneously. The pose estimation iteration process was based on para-perspective projection model,where the coordination of the projection of the object gravity center used as the initial parameter of iteration is not needed. The iteration process of topological determination was transformed into a solution of assignment problem. Each topological determination can obtain a better pose estimation in every pose estimation iteration loop. The results of multiple poses experiment and precision comparison experiment prove that the algorithm is qualified for the high precision pose estimation of 3D object with large scale motion, with the root mean square error 0.272° in the range of-120°-120°.
引文
[1]Sun C,Sun P,Wang P.An improvement of pose measurement method using global control points calibration[J].Plos One,2015,10(7):e0133905.
[2]Chen Haixin,Gu Guohua,Qian Weixian,et al.A Fast method for moving target detection on a moving carema[J].Acta Photonica Sinica,2014,43(3):133-138.(in Chinese)陈海欣,顾国华,钱惟贤,等.一种快速的单目移动相机下运动目标检测方法[J].光子学报,2014,43(3):133-138.
[3]Neumayr R,Zsombor-Murray P,O′Leary P,et al.Precise pose measurement with single camera calibration for planar parallel manipulators[J].Transactions-Canadian Society for Mechanical Engineering,2011,35(35):201-213.
[4]Zeng Zhankui,Gu Qiangwei,Cao Xibin,et al.Relative pose monocular vision determination of spacecraft using orthogonal Procrustes analysis[J].Infrared and Laser Engineering,2015,44(S1):113-118.(in Chinese)曾占魁,谷蔷薇,曹喜滨,等.基于正交Procrustes分析的航天器单目视觉相对位姿确定方法[J].红外与激光工程,2015,44(S1):113-118.
[5]Wu Fuchao,Hu Zhanyi.A linear method for the Pn P problem[J].Journal of Software,2003,14(3):682-688.(in Chinese)吴福朝,胡占义.Pn P问题的线性求解算法[J].软件学报,2003,14(3):682-688.
[6]Duan Fuqing,Wu Fuchao,Hu Zhanyi.Camera calibration and 3D reconstruction using parallelism constraint[J].Journal of Software,2007,18(6):1350-1360.(in Chinese)段福庆,吴福朝,胡占义.基于平行性约束的摄像机标定与3D重构[J].软件学报,2007,18(6):1350-1360.
[7]Wang Peng,Sun Changku,Zhang Zimiao.Linear pose estimation with a monocular vision system[J].Chinese Journal of Scientific Instrument,2011,32(5):1126-1131.(in Chinese)王鹏,孙长库,张子淼.单目视觉位姿测量的线性求解[J].仪器仪表学报,2011,32(5):1126-1131.
[8]Zhang Zimiao,Wang Peng,Sun Changku.Position andorientation measurement method with monocular vision and digital simulation[J].Journal of Tianjin University,2011,44(5):440-444.(in Chinese)张子淼,王鹏,孙长库.单目视觉位姿测量方法及数字仿真[J].天津大学学报,2011,44(5):440-444.
[9]Liu Chang,Zhu Feng,Xia Renbo.Monocular pose determination from coplanar two points and one line features[J].Application Research of Computers,2012,29(8):3145-3147.(in Chinese)刘昶,朱枫,夏仁波.基于共面二点一线特征的单目视觉定位[J].计算机应用研究,2012,29(8):3145-3147.
[10]Dementhon D F,Davis L S.Model-based object pose in 25lines of code[J].International Journal of Computer Vision,1995,15(1-2):123-141.
[11]Horaud R,Dornaika F,Lamiroy B.Object pose:the link between weak perspective,paraperspective,and full perspective[J].International Journal of Computer Vision,1997,22(2):173-189.
[12]Xia Junjian,Xu Xiaoquan,Xiong Jiulong.Iterative pose estimation using paraperspective camera model[J].Optics and Precision Engineering,2012,20(6):1342-1349.(in Chinese)夏军菅,徐小泉,熊九龙.利用平行透视投影模型的位姿迭代估计[J].光学精密工程,2012,20(6):1342-1349.
[13]Zhou Jing,Gao Yinhan,Liu Changying,et al.Attitude caculation of single camera visual system based on an adaptive algorithm[J].Optics and Precision Engineering,2012,20(17):2796-2803.(in Chinese)周婧,高印寒,刘长英,等.基于自适应算法的单目视觉系统的姿态解算[J].光学精密工程,2012,20(17):2796-2803.
[14]David P,Dementhon D,Duraiswami R,et al.Soft POSIT:simultaneous pose and correspondence determination[J].International Journal of Computer Vision,2002,59(3):424-431.
[15]Ugurdag H F,Gren S,Canbay F.Gravitational pose estimation[J].Computers&Electrical Engineering,2010,36(6):1165-1180.
[16]Kriener,Charles R.A comparison and evaluation of three different pose estimation algorithms in detecting low texture manufactured objects[J].Dissertations&Theses-Gradworks,2011(3):108-111.
[17]Gold S,Rangarajan A,Lu C P,et al.New algorithms for2D and 3D point matching:pose estimation and correspondence[J].Pattern Recognition,1999,31(8):1019-1031.
[18]Knight P A.The sinkhorn-knopp algorithm:convergence and applications[J].Siam Journal on Matrix Analysis&Applications,2007,30(1):261-275.
[19]Zhang Z.A flexible new technique for camera calibration[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2000,22(11):1330-1334.
[2]Chen Haixin,Gu Guohua,Qian Weixian,et al.A Fast method for moving target detection on a moving carema[J].Acta Photonica Sinica,2014,43(3):133-138.(in Chinese)陈海欣,顾国华,钱惟贤,等.一种快速的单目移动相机下运动目标检测方法[J].光子学报,2014,43(3):133-138.
[3]Neumayr R,Zsombor-Murray P,O′Leary P,et al.Precise pose measurement with single camera calibration for planar parallel manipulators[J].Transactions-Canadian Society for Mechanical Engineering,2011,35(35):201-213.
[4]Zeng Zhankui,Gu Qiangwei,Cao Xibin,et al.Relative pose monocular vision determination of spacecraft using orthogonal Procrustes analysis[J].Infrared and Laser Engineering,2015,44(S1):113-118.(in Chinese)曾占魁,谷蔷薇,曹喜滨,等.基于正交Procrustes分析的航天器单目视觉相对位姿确定方法[J].红外与激光工程,2015,44(S1):113-118.
[5]Wu Fuchao,Hu Zhanyi.A linear method for the Pn P problem[J].Journal of Software,2003,14(3):682-688.(in Chinese)吴福朝,胡占义.Pn P问题的线性求解算法[J].软件学报,2003,14(3):682-688.
[6]Duan Fuqing,Wu Fuchao,Hu Zhanyi.Camera calibration and 3D reconstruction using parallelism constraint[J].Journal of Software,2007,18(6):1350-1360.(in Chinese)段福庆,吴福朝,胡占义.基于平行性约束的摄像机标定与3D重构[J].软件学报,2007,18(6):1350-1360.
[7]Wang Peng,Sun Changku,Zhang Zimiao.Linear pose estimation with a monocular vision system[J].Chinese Journal of Scientific Instrument,2011,32(5):1126-1131.(in Chinese)王鹏,孙长库,张子淼.单目视觉位姿测量的线性求解[J].仪器仪表学报,2011,32(5):1126-1131.
[8]Zhang Zimiao,Wang Peng,Sun Changku.Position andorientation measurement method with monocular vision and digital simulation[J].Journal of Tianjin University,2011,44(5):440-444.(in Chinese)张子淼,王鹏,孙长库.单目视觉位姿测量方法及数字仿真[J].天津大学学报,2011,44(5):440-444.
[9]Liu Chang,Zhu Feng,Xia Renbo.Monocular pose determination from coplanar two points and one line features[J].Application Research of Computers,2012,29(8):3145-3147.(in Chinese)刘昶,朱枫,夏仁波.基于共面二点一线特征的单目视觉定位[J].计算机应用研究,2012,29(8):3145-3147.
[10]Dementhon D F,Davis L S.Model-based object pose in 25lines of code[J].International Journal of Computer Vision,1995,15(1-2):123-141.
[11]Horaud R,Dornaika F,Lamiroy B.Object pose:the link between weak perspective,paraperspective,and full perspective[J].International Journal of Computer Vision,1997,22(2):173-189.
[12]Xia Junjian,Xu Xiaoquan,Xiong Jiulong.Iterative pose estimation using paraperspective camera model[J].Optics and Precision Engineering,2012,20(6):1342-1349.(in Chinese)夏军菅,徐小泉,熊九龙.利用平行透视投影模型的位姿迭代估计[J].光学精密工程,2012,20(6):1342-1349.
[13]Zhou Jing,Gao Yinhan,Liu Changying,et al.Attitude caculation of single camera visual system based on an adaptive algorithm[J].Optics and Precision Engineering,2012,20(17):2796-2803.(in Chinese)周婧,高印寒,刘长英,等.基于自适应算法的单目视觉系统的姿态解算[J].光学精密工程,2012,20(17):2796-2803.
[14]David P,Dementhon D,Duraiswami R,et al.Soft POSIT:simultaneous pose and correspondence determination[J].International Journal of Computer Vision,2002,59(3):424-431.
[15]Ugurdag H F,Gren S,Canbay F.Gravitational pose estimation[J].Computers&Electrical Engineering,2010,36(6):1165-1180.
[16]Kriener,Charles R.A comparison and evaluation of three different pose estimation algorithms in detecting low texture manufactured objects[J].Dissertations&Theses-Gradworks,2011(3):108-111.
[17]Gold S,Rangarajan A,Lu C P,et al.New algorithms for2D and 3D point matching:pose estimation and correspondence[J].Pattern Recognition,1999,31(8):1019-1031.
[18]Knight P A.The sinkhorn-knopp algorithm:convergence and applications[J].Siam Journal on Matrix Analysis&Applications,2007,30(1):261-275.
[19]Zhang Z.A flexible new technique for camera calibration[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2000,22(11):1330-1334.