A Perspective Theory for Motion and Shape Estimation in Machine Vision

作者: B. K. Ghosh , E. P. Loucks

DOI: 10.1137/S0363012993246853

关键词:

摘要: In this paper, we consider the problem of motion and shape estimation a moving body with aid monocular camera. We show that reduces to specific parameter perspective dynamical system. Surprisingly, above reduction is independent whether data measured brightness pattern which object produces on image plane or observed are points, lines, curves produced as result discontinuities in pattern. Many cases have been analyzed paper. These include fairly complete analysis planar textured surface undergoing rigid flow an affine flow. two for orthographic, pseudo-orthographic, image-centered projections. The basic procedure introduced subdivide into modules, one "spatial averaging" other "time averaging." carried out new "realization theory systems" systems described discrete time continuous time. Finally, much our has substantiated by computer simulation algorithms developed order explicitly compute parameters. Detailed would answer realizability question subject future research.

参考文章(5)
A. N. Netravali, T. S. Huang, A. S. Krishnakumar, R. J. Holt, Algebraic methods in 3‐d motion estimation from two‐view point correspondences International Journal of Imaging Systems and Technology. ,vol. 1, pp. 78- 99 ,(1989) , 10.1002/IMA.1850010110
Minas E. Spetsakis, John (Yiannis) Aloimonos, Structure from motion using line correspondences International Journal of Computer Vision. ,vol. 4, pp. 171- 183 ,(1990) , 10.1007/BF00054994
John Canny, A Computational Approach to Edge Detection IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. PAMI-8, pp. 679- 698 ,(1986) , 10.1109/TPAMI.1986.4767851
David St. George Vernon, Robot Vision ,(1986)