Vision-based road following for an autonomous land vehicle
Abstract
An approach to visual navigation developed under the Autonomous Land Vehicle project is presented. The approach includes a bootstrap phase and a feed-forward phase, with visual processing in each phase executed in three modules: an image processing module, a geometry module, and a rule-based reasoning module. Process control for road following is described, with particular attention given to the image processing algorithms for the bootstrap phase. Guidelines for future work are briefly discussed.
- Publication:
-
5th Computers in Aerospace Conference
- Pub Date:
- 1985
- Bibcode:
- 1985coae.conf..453W
- Keywords:
-
- Artificial Intelligence;
- Autonomous Navigation;
- Computer Vision;
- Image Processing;
- Terrain Following Aircraft;
- Visual Flight;
- Architecture (Computers);
- Digital Navigation;
- Expert Systems;
- Feedforward Control;
- Surface Vehicles;
- Engineering (General)