Real time image processing for autonomous guided road vehicles makes great demands on the computation power installed in the car. From the image sequence, delivered by a standard video camera, road course, position and motion of potential obstacles as well as traffic signs have to be extracted. Since 1985 Daimler Benz Research is working on autonomous road vehicles guided by video cameras. In 1989 the authors decided to use transputers as a flexible and powerful development platform for the real time image processing and control tasks. Since August 1991 the T8 systems are on public roads, when the authors performed the first long distance trip of about 130 km on a highway close to Stuttgart. Meanwhile the authors gained experience from more than 2000 km autonomous driving, including a 600 km trip from Stuttgart to Berlin. This paper presents the state of the art of the work on lane keeping, obstacle detection, traffic sign recognition and development tools for parallelization. Besides the great challenge of an autonomous vehicle various spin offs have been derived from the basic techniques that can help the driver in standard as well as in critical situations. The techniques developed as well as possible applications are described in this contribution.
Transputers on the road
Transputer auf der Straße
1993
16 Seiten, 10 Bilder, 11 Quellen
Conference paper
English
Onboard Image Processing With Transputers
British Library Conference Proceedings | 1994
|Visualization and View Simulation Based on Transputers
Online Contents | 1993
|Analysing Road Traffic Movement Patterns By Image Analysis Using An Array of Transputers
British Library Conference Proceedings | 1996
|Visualization and view simulation based on transputers
AIAA | 1993
|TRANSPUTERS AND THEIR USE IN ON-BOARD COMPUTERS
Online Contents | 1996
|