Abstract

A significant amount of work has been devoted to the estimation of the motion parameters of a rigid body based on a sequence of images. As discussed in [1], the problem of noisy point correspondences (match points) in the estimation of 3-D motion from 2-D images is as yet unresolved. In [2], a linear filtering approach is introduced for the case of a longer sequence of noisy images of a rigid body whose motion is restricted to a plane. In this summary, a nonlinear filtering approach is presented. By using a sequence of 10 to 30 images and one or more match points, an Iterated Extended Kalman (IEKF) Filter can be used to estimate motion parameters in the presence of significant measurement noise. The ultimate goal of this approach is the estimation of 3-D motion parameters from a sequence of noisy 2-D images. The feasibility of the approach is demonstrated here for the case of a noisy 1-D image of a 2-D rigid body.

© 1985 Optical Society of America