Real-Time Vegetation Discrimination

PDF

Authors
  1. Tarrant, S.
  2. Piercey, G.
  3. Hart, D.
  4. McGuire, P.
Corporate Authors
Defence R&D Canada - Suffield, Ralston ALTA (CAN);C-Core - St. John's, NL (CAN)
Abstract
The major objectives for this project involve the development of a vegetation discrimination algorithm and system for use on an unmanned ground vehicle (UGV). Currently, UGV's have great difficulty in distinguishing between true obstacles such as rocks, berms and manmade structures, and terrain cover such as grass, foliage, dry vegetation and bark. Consequently, terrain with vegetation is usually modeled as obstacles even though a human driver would easily traverse it. Though falsely identifying vegetation as obstacles does not imperil a UGV, it does impose artificial and unnecessary limits on the vehicle.
Keywords
Software;Unmanned ground vehicles;UGV;Terrain cover;FD-1665 multispectral FireWire camera;Discrimination algorithm;Vegetation detection
Report Number
DRDC-SUFFIELD-CR-2009-101 — Contractor Report
Date of publication
01 Mar 2009
Number of Pages
54
DSTKIM No
CA033698
CANDIS No
533061
Format(s):
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: