Direct Obstacle Detection and Motion from Spatio-Temporal Derivatives

Pär Fornland

Proc. 6th International Conference on Computer Analysis of Images and Patterns (Prague, Chech Republic), September 1995.

Abstract

Autonomous vehicles need a means of detecting obstructions on its path, to avoid collision. In this paper, a novel approach to obstacle detection is presented. A camera moves on a visible ground plane with the optical axis parallel to the ground. Camera motion parameters are linearly related to first order spatio-temporal derivatives of the taken image sequence; image flow is not neenada MotionCVAP/ is robustly estimated using RANSAC. An error measure for each image point corresponds to the likelihood of an obstacle in that point.

Keywords: obstacle detection, motion estimation, optical flow, computer vision, autonomous vehicles

CAIP'95: (Postscript 401k)


Pär Fornland <par@bion.kth.se>