Direct Determination of General Motion from Normal Flows

Tak-Wai Hui and Ronald Chung


General Motion, Optical Flows, Normal Flows, Direct Method


Determining object motion from video data is important for computer vision and various robotics tasks including visual servoing and autonomous navigation. The problem is especially difficult if the motion in space can be a general one. The difficulty lies in that the flow pattern directly observable in the video is generally not the full flow field induced by the motion, but only partial information of it, which is known as the normal flow field. A few methods, collectively referred to as the direct methods, have been proposed to determine the spatial motion from merely the normal flow field without ever interpolating the full flows. Yet they either put restrictions on the nature of the motion itself, or they are operable only under rather limiting conditions. This work proposes a new direct method to determine motion using all observable normal flows. The constraint presents itself as a system of linear inequalities to bind the motion parameters. An iterative process in a double coarse-to-fine framework on the motion parameter space is used to exploit the constraint. Experimental results on benchmarking data are presented.

Important Links:

Go Back