Vision based reconstruction and pose estimation for spacecraft with axisymmetric structure

Abstract Spacecraft with axisymmetric structure is simplified to an effective model based on its cross sections, and a novel reconstruction and relative pose estimation algorithm is presented in this paper. Firstly, features (i.e. line or ellipse) on the image projected by cross sections of axisymmetric structure are obtained based on contour constraint and center constraint. Secondly, the simplified model is reconstructed by calculating the normalized size and distances of cross-section features. In addition, a kind of singular projection is analyzed when a ellipse projected by circular cross section degenerates to a straight line on the image plane. Finally, relative pose between target and camera is calculated by these features. The method is tested on numerous synthetic images, and experimental results indicate that the method perform well in target recognition and reconstruction.

[1]  Yinqiang Zheng,et al.  Camera calibration using two concentric circles: linear approach , 2009 .

[2]  Kostas Daniilidis,et al.  Linear Pose Estimation from Points or Lines , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  Hui Zhang,et al.  The common self-polar triangle of concentric circles and its application to camera calibration , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Pierre Gurdjos,et al.  Euclidean Structure from N geq 2 Parallel Circles: Theory and Algorithms , 2006, ECCV.

[5]  Jun-Sik Kim,et al.  Geometric and algebraic constraints of projected concentric circles and their applications to camera calibration , 2005, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Long Quan,et al.  Detection of concentric circles for camera calibration , 2005, Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1.

[7]  叶东 Ye Dong,et al.  Binocular Vision Method of Measuring Pose Based on Perpendicular Lines , 2014 .

[8]  Philip David,et al.  SoftPOSIT: Simultaneous Pose and Correspondence Determination , 2002, International Journal of Computer Vision.

[9]  B. Benhabib,et al.  An analytical method for the 3D-location estimation of circular features for an active-vision system , 1990, 1990 IEEE International Conference on Systems, Man, and Cybernetics Conference Proceedings.

[10]  V. Frémont,et al.  DIRECT CAMERA CALIBRATION USING TWO CONCENTRIC CIRCLES FROM A SINGLE VIEW , 2002 .

[11]  Dong Ye,et al.  Multiple circle recognition and pose estimation for aerospace application , 2017 .

[12]  Shourui Yang,et al.  An improved method for location of concentric circles in vision measurement , 2017 .

[13]  Roland Siegwart,et al.  Stereo-Based Ego-Motion Estimation Using Pixel Tracking and Iterative Closest Point , 2006, Fourth IEEE International Conference on Computer Vision Systems (ICVS'06).

[14]  Songde Ma,et al.  Conics-based stereo, motion estimation, and pose determination , 1993, International Journal of Computer Vision.

[15]  Shengyong Chen,et al.  Concentric-circle-based camera calibration , 2012 .

[16]  Shuanghe Yu,et al.  The location and identification of concentric circles in automatic camera calibration , 2013 .

[17]  Reinhard Koch,et al.  Robust and Efficient Pose Estimation from Line Correspondences , 2012, ACCV.