三维目标位姿跟踪与模型修正

doi:10.11947/j.AGCS.2018.20170626

摘要/Abstract

摘要： 本文是机器视觉参量下的三维数字摄影测量智能构像基础工作之一：成像系统位置姿态自动跟踪与精密修正，属于摄影测量与机器视觉、数字图像处理等学科交叉的摄像测量领域。针对基于目标3D模型的位姿跟踪问题开展研究，对相关研究的现状进行梳理，并提出系列位姿跟踪与模型修正方法。在完全已知目标3D精确模型的情况下，对于包含丰富直线特征的特殊目标，提出基于直线模型的目标位姿跟踪方法，实现了目标位姿参数的精确跟踪；为处理更为一般目标，利用目标的3D边缘模型，提出法向距离迭代加权最小二乘位姿估计方法及距离图迭代最小二乘位姿跟踪方法。当目标3D直线模型参数不准确时，结合光束法平差思想，提出一种针对序列图像的基于3D直线模型同时位姿跟踪与模型修正方法，联合优化求解目标位姿参数及3D直线模型参数，在模拟空间卫星目标位姿测量的仿真试验中，模型直线朝向、位置误差及目标位姿平均角度、平均位置误差分别为0.3°、3.5 mm及0.12°、20.1 mm。针对包含丰富直线特征的目标，在其3D直线模型完全未知的情况下，提出基于序列图像直线对应的目标结构重建与位姿跟踪方法，利用序列图像信息，在SFM框架下同时优化求解目标直线模型参数及位姿参数，仿真试验条件下，重建模型直线朝向、位置误差及位姿参数平均角度、平均位置误差分别约为0.4°、7.5 mm及0.16°、23.5 mm。

关键词: 摄像测量, 3D模型, 位姿跟踪, 光束法平差, 三维重建

Abstract: This paper tackles imaging system pose tracking and model refinement,one of the fundamental work for 3D photogrammetry.The researches belong to the videometrics,an interdiscipline which combines computer vision,digital image processing,photogrammetry and optical measurement.Related works are summarized briefly in this paper.We study the problems of pose tracking for target with 3D model.For the target with accurate 3D model,line model based pose tracking methods are proposed for target with rich line features.Experimental results indicate that the proposed methods track the target pose accurately.Normal distance iterative reweighted least squares and distance image iterative least squares methods are proposed to process more general targets.This paper adopts bound adjustment to tackle pose tracking in image sequence for target with inaccurate 3D line model.The proposed method optimizes model line parameters and pose parameters simultaneously.The model line orientation,position and mean angle error,mean position error of pose are 0.3°,3.5 mm and 0.12°,20.1 mm in simulation experiments of satellite pose tracking.Line features are used to track target pose with unknown 3D model through image sequence.The model line parameters and pose parameters are optimized under the framework of SFM.In simulation experiments,the reconstructed line orientation,position error and mean angle error,mean position error of pose are 0.4°,7.5 mm and 0.16°,23.5 mm.

Key words: videometrics, 3D model, pose tracking, bound adjustment, reconstruction

中图分类号:

TP391

尚洋, 孙晓亮, 张跃强, 李由, 于起峰. 三维目标位姿跟踪与模型修正[J]. 测绘学报, 2018, 47(6): 799-808.

SHANG Yang, SUN Xiaoliang, ZHANG Yueqiang, LI You, YU Qifeng. Research on 3D Target Pose Tracking and Modeling[J]. Acta Geodaetica et Cartographica Sinica, 2018, 47(6): 799-808.

参考文献

[1] 于起峰, 尚洋. 摄像测量学原理与应用研究[M]. 北京:科学出版社, 2009. YU Qifeng, SHANG Yang. Videometrics:Principles and Researches[M]. Beijing:Science Press, 2009.
[2] 尚洋. 基于视觉的空间目标位置姿态测量方法研究[D]. 长沙:国防科技大学, 2006. SHANG Yang. Researches on Vision-based Pose Measurements for Space Target[D]. Changsha:National University of Defense Technology, 2006.
[3] LOWE D G. Distinctive Image Features from Scale-invariant Keypoints[J]. International Journal of Computer Vision, 2004, 60(2):91-110.
[4] RUBLEE E, RABAUD V, KONOLIGE K, et al. ORB:An Efficient Alternative to SIFT or SURF[C]//IEEE International Conference on Computer Vision. Barcelona, Spain:IEEE, 2011:2564-2571.
[5] FISCHLER M A, BOLLES R C. Random Sample Consensus:A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography[J]. Communications of the ACM, 1981, 24(6):381-395.
[6] DEMENTHON D F, DAVIS L S. Model-based Object Pose in 25 Lines of Code[J]. International Journal of Computer Vision, 1995, 15(1-2):123-141.
[7] LU C P, HAGER G D, MJOLSNESS E. Fast and Globally Convergent Pose Estimation from Video Images[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(6):610-622.
[8] DHOME M, RICHETIN M, LAPRESTÉ J T, et al. Determination of the Attitude of 3D Objects from a Single Perspective View[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1989, 11(12):1265-1278.
[9] CHEN H H. Pose Determination from Line-to-plane Correspondences:Existence Condition and Closed-form Solutions[C]//IEEE Third International Conference on Computer Vision. Osaka, Japan, Japan:IEEE, 1990:374-378.
[10] ZHANG Lilian, XU Chi, LEE K M, et al. Robust and Efficient Pose Estimation from Line Correspondences[M]//LEE K M, MATSUSHITA Y, REHG J M, et al. Computer Vision-ACCV 2012. Berlin, Heidelberg:Springer, 2012:217-230.
[11] 李鑫, 张跃强, 刘进博, 等. 基于直线段对应的相机位姿估计直接最小二乘法[J]. 光学学报, 2015, 35(6):615003. LI Xin, ZHANG Yueqiang, LIU Jinbo, et al. A Direct Least Squares Method for Camera Pose Estimation Based on Straight Line Segment Correspondences[J]. Acta Optica Sinica, 2015, 35(6):615003.
[12] LIU Y, HUANG T S, FAUGERAS O D. Determination of Camera Location from 2D to 3D Line and Point Correspondences[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990, 12(1):28-37.
[13] CHRISTY S, HORAUD R. Iterative Pose Computation from Line Correspondences[J]. Computer Vision and Image Understanding, 1999, 73(1):137-144.
[14] DAVID P, DEMENTHON D, DURAISWAMI R, et al. Simultaneous Pose and Correspondence Determination Using Line Features[C]//IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Madison, WI:IEEE, 2003:424-431.
[15] 王竞雪, 宋伟东, 王伟玺. 同名点及高程平面约束的航空影像直线匹配算法[J]. 测绘学报, 2016, 45(1):87-95. DOI:10.11947/j.AGCS.2016.20140527. WANG Jingxue, SONG Weidong, WANG Weixi. Line Matching Algorithm for Aerial Image Based on Corresponding Points and Z-plane Constraints[J]. Acta Geodaetica et Cartographica Sinica, 2016, 45(1):87-95. DOI:10.11947/j.AGCS.2016.20140527.
[16] 曹金山, 龚健雅, 袁修孝. 直线特征约束的高分辨率卫星影像区域网平差方法[J]. 测绘学报, 2015, 44(10):1100-1107. DOI:10.11947/j.AGCS.2015.20150023. CAO Jinshan, GONG Jianya, YUAN Xiuxiao. A Block Adjustment Method of High-resolution Satellite Imagery with Straight Line Constraints[J]. Acta Geodaetica et Cartographica Sinica, 2015, 44(10):1100-1107. DOI:10.11947/j.AGCS.2015.20150023.
[17] HARRIS C, STENNETT C. RAPID-A Video Rate Object Tracker[C]//Proceedings of the British Machine Vision Conference.[s.l.]:BMVA Press, 1990:73-78.
[18] COMPORT A I, MARCHAND E, PRESSIGOUT M, et al. Real-time Markerless Tracking for Augmented Reality:the Virtual Visual Servoing Framework[J]. IEEE Transactions on Visualization and Computer Graphics, 2006, 12(4):615-628.
[19] COMPORT A I, KRAGIC D, MARCHAND E, et al. Robust Real-time Visual Tracking:Comparison, Theoretical Analysis and Performance Evaluation[C]//IEEE International Conference on Robotics and Automation. Barcelona, Spain:IEEE, 2005:2841-2846.
[20] 张跃强. 基于直线特征的空间非合作目标位姿视觉测量方法研究[D]. 长沙:国防科技大学, 2016. ZHANG Yueqiang. Research on Vision Based Pose Measurement Methods for Space Uncooperative Objects Using Line Features[D]. Changsha:National University of Defense Technology, 2006.
[21] CHOI C, Christensen H I. Real-Time 3D Model-based Tracking Using Edge and Keypoint Features for Robotic Manipulation[C]//IEEE International Conference on Robotics and Automation. Anchorage, AK:IEEE, 2010:4048-4055.
[22] BROX T, ROSENHAHN B, GALL J, et al. Combined Region and Motion-Based 3D Tracking of Rigid and Articulated Objects[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(3):402-415.
[23] DRUMMOND T, CIPOLLA R. Real-time Tracking of Complex Structures with On-line Camera Calibration[C]//Proceedings of British Machine Vision Conference. Nottingham:BMVC, 1999:574-583.
[24] ARMSTRONG M, ZISSERMAN A. Robust Object Tracking[C]//Proceedings of Second Asian Conference on Computer Vision. Singapore:[s.n.], 1995:58-62.
[25] YOON Y, KOSAKA A, KAK A C. A New Kalman-filter-based Framework for Fast and Accurate Visual Tracking of Rigid Objects[J]. IEEE Transactions on Robotics, 2008, 24(5):1238-1251.
[26] CHOI C, CHRISTENSEN H I. Robust 3D Visual Tracking Using Particle Filtering on the SE(3) Group[C]//IEEE International Conference on Robotics and Automation. Shanghai, China:IEEE, 2011:4384-4390.
[27] MUNDY J L. Object Recognition in the Geometric Era:A Retrospective[M]//PONCE J, HEBERT M, SCHMID C, et al. Toward Category-level Object Recognition. Berlin, Heidelberg:Springer, 2010:3-28.
[28] LIM J J, PIRSIAVASH H, TORRALBA A. Parsing IKEA Objects:Fine Pose Estimation[C]//IEEE International Conference on Computer Vision. Sydney, NSW, Australia:IEEE, 2014:2992-2999.
[29] CHOY C B, STARK M, CORBETT-DAVIES S, et al. Enriching Object Detection with 2D-3D Registration and Continuous Viewpoint Estimation[C]//IEEE Conference on Computer Vision and Pattern Recognition. Boston, MA:IEEE, 2015:2512-2520.
[30] FIDLER S, DICKINSON S J, URTASUN R. 3D Object Detection and Viewpoint Estimation with a Deformable 3D Cuboid Model[C]//International Conference on Neural Information Processing Systems. Lake Tahoe, Nevada:ACM, 2012:611-619.
[31] MOTTAGHI R, XIANG Yu, SAVARESE S. A Coarse-to-fine Model for 3D Pose Estimation and Sub-Category Recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition. Boston, MA:IEEE, 2015:418-426.
[32] WOHLHART P, LEPETIT V. Learning Descriptors for Object Recognition and 3D Pose Estimation[C]//IEEE Conference on Computer Vision and Pattern Recognition. Boston, MA:IEEE, 2015:3109-3118.
[33] CRIVELLARO A, RAD M, VERDIE Y, et al. A Novel Representation of Parts for Accurate 3D Object Detection and Tracking in Monocular Images[C]//IEEE International Conference on Computer Vision. Santiago, Chile:IEEE, 2015:4391-4399.
[34] DISSANAYAKE M W M G, NEWMAN P, CLARK S, et al. A Solution to the Simultaneous Localization and Map Building (Slam) Problem[J]. IEEE Transactions on Robotics and Automation, 2001, 17(3):229-241.
[35] DAVISON A J, REID I D, MOLTON N D, et al. MonoSLAM:Real-time Single Camera SLAM[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 29(6):1052-1067.
[36] KONOLIGE K, AGRAWAL M. FrameSLAM:From Bundle Adjustment to Real-time Visual Mapping[J]. IEEE Transactions on Robotics, 2008, 24(5):1066-1077.
[37] AUGENSTEIN S. Monocular Pose and Shape Estimation of Moving Targets, for Autonomous Rendezvous and Docking[D]. Stanford, CA:Stanford University, 2011.
[38] 李由. 基于轮廓和边缘的空间非合作目标视觉跟踪[D]. 长沙:国防科技大学, 2013. LI You. Contour and Edge-based Visual Tracking of Non-cooperative Space Targets[D]. Changsha:National University of Defense Technology, 2013.
[39] VON GIOI R G, JAKUBOWICZ J, MOREL J M, et al. LSD:A Fast Line Segment Detector with a False Detection Control[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(4):722-732.