New method for camera pose estimation based on line correspondence

The key technology of the measuring method of pose shaking of unstable platform based on CCD lies in the camera pose estimation. A new method for camera pose estimation based on line correspondence is put forward in this paper. First, two appropriate lines are selected in the space to establish a local coordinate system, the remaining lines form three-line sets with these two lines respectively and then the angle depth of the lines of the local coordinate system can be obtained by using the linearity of interpretation plane; then the invariance of included angle between spatial coordinate system and camera coordinate system during rigid transformation is used to establish a constraint equation set of angle depth of three lines similar to traditional P3P, and the corresponding angle depth of the third line in the set can be obtained linearly by variable substitution; finally, the rotation matrix can be expressed with quaternion and the pose parameters of camera in the space can be calculated by singular value decomposition (SVD). The static experiment and dynamic experiment verify the accuracy and stability of the algorithm and that the camera pose shocking is within 1° and the 3σ error is superior to 0.3′. The method in this paper can be directly used in the micro-vibration measurement of experimental platform and robot navigation and other fields and is of important theoretical significance and engineering application prospect.

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

[2]  Qifeng Yu,et al.  Relay camera videometrics based conversion method for unstable platform to static reference , 2011 .

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

[4]  王. W. Fang,et al.  Analysis and correction for influence of vehicle platform deformation on measuring errors , 2011 .

[5]  Qifeng Yu,et al.  Robust camera pose estimation from unknown or known line correspondences. , 2012, Applied optics.

[6]  Nassir Navab,et al.  Yet another method for pose estimation: A probabilistic approach using points, lines, and cylinders , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[7]  Radu Horaud,et al.  Object pose from 2-D to 3-D point and line correspondences , 1995, International Journal of Computer Vision.

[8]  Luc Van Gool,et al.  Generalised Linear Pose Estimation , 2007, BMVC.

[9]  Stéphane Christy,et al.  Fast and Reliable Object Pose Estimation from Line Correspondences , 1997, CAIP.

[10]  Olivier D. Faugeras,et al.  Monocular pose determination from lines: critical sets and maximum number of solutions , 1993, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[11]  Robert M. Haralick,et al.  An Integrated Linear Technique for Pose Estimation from Different Geometric Features , 1999, Int. J. Pattern Recognit. Artif. Intell..

[12]  Qifeng Yu,et al.  Analytical method for camera calibration from a single image with four coplanar control lines. , 2004, Applied optics.

[13]  Homer H. Chen Pose Determination from Line-to-Plane Correspondences: Existence Condition and Closed-Form Solutions , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[14]  Olivier D. Faugeras,et al.  Determination of Camera Location from 2-D to 3-D Line and Point Correspondences , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[15]  Fei Wang,et al.  Linear Pose Estimation Algorithm Based on Quaternion , 2011, ICIC.

[16]  Michel Dhome,et al.  Determination of the Attitude of 3D Objects from a Single Perspective View , 1989, IEEE Trans. Pattern Anal. Mach. Intell..