New Results in Stereo-Based Automatic Vehicle Guidance

Joseph Weber , Quang-Tuan Luong, Dieter Koller, and Jitendra Malik

In Proc. of the Intelligent Vehicles Symposium 1995, September 25-26, 1994, Detroit, USA

Abstract

We present new results on vision based longitudinal and lateral vehicle control. As described in our previous work presented at Intelligent Vehicles 94, the novel feature of this approach is the use of binocular vision. We integrate two modules consisting of a domain-specific, efficient binocular stereo algorithm, and a lane marker detection algorithm, and show that the integration results in a improved performance for each of the modules.

Longitudinal control is supported by detecting and measuring the distances to leading vehicles using binocular stereo. The knowledge of the camera geometry with respect to the locally planar road is used to map the images of the road plane in the two camera views into alignment. This allows us to separate image features into those lying in the road plane, e.g. lane markers, and those due to other objects which are above the ground plane and thus a potential hazard. In contrast with previous work using vision for vehicle control, by using stereo information we can cope with the difficulties arising from occlusion of lane markers by other vehicles.

Features detected above the ground plane are labeled by a simple vehicle model. Position and relative velocity information is then tracked with time using new information and dynamic constraints. The result is a dynamic map of the position and relative velocity of visible obstacles. From this the appropriate driving control can be derived.

The detection and measurement of the lane markers provides us with the positional parameters and the road curvature which are needed for lateral vehicle control. Moreover, this information is also used to update the camera geometry with respect to the road, therefore allowing us to cope with the problem of vibrations and road inclination to obtain consistent results from binocular stereo. A simple second order system is shown to provide accurate positional information for camera orientation update.

Results of sensor output from video input will be presented, as well as the results of trial runs of a simple control module using the visual data on our test vehicle.

The document is available online in application/postscript (810976 Bytes)


Last modified on Tuesday, November 20, 1996, Dieter Koller (koller@vision.caltech.edu)