Range imagery from a laser range imager developed at ERIM to support the DARPA ALV program provides a robust source of information to direct autonomous vehicle movement. Forward looking range imagery is processed into plan view map information of the surrounding environment for guidance and identifying obstacles. Intensity data from the sensor also provides a correlated object reflectivity image. The image processing required to transform the forward looking range data into top down format, and to subsequently find the vehicle path is completed at a nominal two second frame rate. If and when image processing speed is comparable to sensor rate (two frames per second), this would support a vehicle speed of approximately 40 miles per hour. This processing rate is possible using the current generation of Cytocomputer ® technology and can be further improved by migrating more of the image processing algorithm from a host VAX to the dedicated Cytocomputer processing system.
Autonomous Vehicle Guidance Using Laser Range Imagery
Mobile Robots II ; 1987 ; Cambridge,CA,United States
Proc. SPIE ; 0852
1987-01-01
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Autonomous Vehicle Guidance Evaluation
NTIS | 1995
|Vehicle Guidance for an Autonomous Vehicle
British Library Conference Proceedings | 1999
|