Recursive Gradient Estimation Using Splines for Navigation of Autonomous Vehicles,
Abstract
Terrain gradient estimation is needed for navigation of an autonomous vehicle in climbing the hills. The in-path and cross-path terrain slopes are estimated from the set of corresponding range slopes. A two-dimensional recursive smoothing algorithm using polynomial splines in the third dimension is developed for this purpose. Approximations are introduced in the sub-optimal system so that the computation time increases only linearly with the size of the two-dimensional data. The successful development of an autonomous vision system for mobile vehicles would be considerable value and importance to defense and related fields. Numerous reports and studies currently recommend artificial intelligence/robotics applications which require autonomous vehicles. Essential to these robotic vehicles is an adequate and efficient computer vision system. A potentially more successful approach, other than TV pictures and photographs, would be to develop a three-dimensional system employing a laser rangefinder.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1985
- Accession Number
- ADP004948
Entities
People
- C. N. Shen
Organizations
- United States Army Armament Research, Development and Engineering Center