THE ACCURACY OF THE COMPUTATION OF OPTICAL-FLOW AND OF THE RECOVERY OF MOTION PARAMETERS

In this paper, the accuracy and the dependence on parameters of a gene ral scheme for the analysis of time-varying image sequences is present ed. The approach, which is based on a modification of a recently propo sed technique for the computation of optical How (Verri et al. [43]), is able to produce vector fields from which it is possible to recover 3-D motion parameters such as time-to-collision and angular velocity. The numerical stability of the computed optical flow and the dependenc e of the recovery of 3-D motion parameters on spatial and temporal fil tering is investigated. By considering optical flows computed on subsa mpled images or along single scanlines, it is also possible to recover 3-D motion parameters from reduced optical flows. An adequate estimat e of time-to-collision can be obtained from sequences of images with s patial resolution reduced to 128 x 128 pixels or from sequences of sin gle scanlines passing near the focus of expansion. The use of Kalman f iltering increases the accuracy and the robustness of the estimation o f motion parameters. The proposed approach seems to be able to provide not only a theoretical background but also practical tools that are a dequate for the analysis of time-varying image sequences.