Recovering 3-D motion parameters from optical flow field using randomized Hough transform

Abstract

This paper addresses the problem of recovering 3-D motion parameters from optical flow fields calculated from two or more consecutive pairs of perspective images. The proposed approach is based on the ideas of Randomized Hough Transform (RHT), i.e., the principles of random sampling of velocity vectors and accumulation of motion parameters. The provided experiments with simulated and real-image data in translational and general motion cases demonstrate that a robust interpretation of 3-D motion parameters can be obtained even in the presence of a significant level of noise.