a method for occlusion recovery is presented based on 3D subspace under orthographic projection. The method mainly relies on the fact that the rows in the matrix including all the image points span the same linear subspace as the rows in the matrix including all the space points. A linear iterative algorithm is hence proposed. The advantages of the proposed method are that the positions of the occlusions are found from all the visible points
and all the images as well as all the image points are treated in the same way. The experiments with both simulated and real data sets show that the re-projection error of the method is approximately 15% higher than the image noise
and the method has good convergence property with small errors.
LIU Zhe, WEI Junhu, ZHAO Jun, et al. Tracking and locating algorithm of 3-dimensional moving targets using 2-dimensional angular measurements from multiple passive sensors[J]. Journal of Xi'an Jiaotong University, 2008, 42(4):462-465.
TOMASI C, KANADE T. Shape and motion from image streams under orthography: a factorization method[J]. International Journal of Computer Vision,1992, 9(2):137-154.
JACOBS D. Linear fitting with missing data: applications to structure from motion and to characterizing intensity images[C]∥Conference on Computer Vision and Pattern Recognition. Los Alamitos, CA, USA: IEEE Computer Society, 1997: 206-212.
MARTINEC D, PAJDLA T. Outlier detection for factorization-based reconstruction from perspective images with occlusions[C]∥Proceedings of the Photogrammetric Computer Vision. Los Alamitos, CA, USA: IEEE Computer Society, 2002: 161-164.
LIU Shigang, PENG Yali, XU Qiuping, et al. Occlusion recovery method based on rank 3[J]. Chinese Journal of Scientific Instrument, 2008, 29(12):2622-2625.
LIU Shigang, WU Chengke, LI Liangfu, et al. Projective reconstruction algorithm with occlusions[J]. Pattern Recognition and Artificial Intelligence, 2007, 20(4): 565-570.
HARTLEY R, ZISSERMAN A. Multiple view geometry in computer vision [M]. Cambridge, UK:Cambridge University Press, 2000.