Title :
Dynamic vision and estimation on spheres
Author_Institution :
Dipt. di Elettronica e Inf., Padova Univ., Italy
Abstract :
In this paper we analyze the simplest kind of estimation problems encountered in dynamic vision, namely tracking an unknown direction from noisy perspective projections on the image plane. We formulate this as an estimation problem on the unit sphere. Assuming a suitable class of probability density functions, we give explicit formulas to compute the steady-state MAP estimate. These formulas look in certain cases like nonlinear recursions of the Kalman filtering type
Keywords :
Bayes methods; computer vision; image reconstruction; optical tracking; probability; recursive estimation; Bayesian perspective estimation; Kalman filtering; computer vision; directional reconstruction; dynamic vision; probability density functions; recursive estimation; tracking; Additive noise; Cameras; Computer vision; Image analysis; Kalman filters; Machine vision; Optical distortion; Optical noise; Probability density function; Steady-state;
Conference_Titel :
Decision and Control, 1997., Proceedings of the 36th IEEE Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
0-7803-4187-2
DOI :
10.1109/CDC.1997.657601
