Inter-frame Registration Using Kalman Filter for Vision Based Navigation

J.C. Yao, E.K. Poh

Abstract:	Inter-frame registration is a crucial component for vision based navigation. Although a wide repertoire of image registration algorithms was developed over the last two decades, the registration of non-texture frames remains a challenging issue. In order to solve the issue, we proposed a new and effective causal scheme for inter-frame registration, which was implemented by a total of 5 Kalman filters. As these Kalman filters naturally incorporate the temporal information contained in a video into the estimation of registration parameters, the algorithm presented in this paper is quite robust and has good performance even for the registration of non-texture frames.
Published in:	Proceedings of the ION 2013 Pacific PNT Meeting April 23 - 25, 2013 Marriott Waikiki Beach Resort & Spa Honolulu, Hawaii
Pages:	374 - 380
Cite this article:	Yao, J.C., Poh, E.K., "Inter-frame Registration Using Kalman Filter for Vision Based Navigation," Proceedings of the ION 2013 Pacific PNT Meeting, Honolulu, Hawaii, April 2013, pp. 374-380.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In