Hindernisdetektion durch Auswertung des optischen Flusses aus Bildfolgen mit Hilfe von Geschwindigkeitsinformation aus Fahrzeugsensoren

In this contribution, we present an approach for the detection of objects in front of a moving vehicle by evaluation of optical flow. It will be shown, how twodimensional optical flow fields can be interpreted to infer information about the three-dimensional environment if the camera is translating on a planar road. Objects are detected by comparing calculated optical flow fields with a model vector field which explicitly expresses the image shift expectation resulting from the camera motion if the environment is free of obstacles. The efficiency of this obstacle detection approach was shown using different algorithms for optical flow vector calculation (6). Until now, we had to determine the camera motion heuristically in order to estimate a model vector field. In this paper we evaluate additional velocity information from vehicle sensors. It will be shown that the obstacle detection procedure using velocity information from vehicle sensors yields results similar to the ones obtained by the heuristic velocity estimation. Results obtained from a real world image sequece to detect obstacles by evaluation of optical flow fields will be presented.