A Barrier Detection Experiment

Abstract

In January 2023, an assessment of research into the use of computer vision for barricade detection was conducted by the Artificial Intelligence for Maneuver and Mobility Essential Research Program. The computer vision algorithm was intended to identify routes that would be traversable except for the presence of movable objects blocking movement along the route. The experiment described here was conducted to assess how the algorithm performed at detecting a variety of barricades across a range of distances and angles. The algorithm, as integrated onto a robotic platform, detected barricades at a distance of approximately 1015 m. The detection was more reliable when the robot was directly facing the barricade than when the robot approached at an angle. The difference in reliability could have been due either to the algorithm or to a camera field of view that did not include the barricade when the robot entered the 15-mdetection range at an angle.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 17, 2023
Accession Number
AD1210688

Entities

People

  • Andre Harrison
  • Craig Lennon
  • L. M. Hernandez
  • Marshal Childrers
  • Seth Ellis

Organizations

  • United States Army Research Laboratory

Tags

Readers

  • Computer Vision.
  • Munitions and Ordnance Engineering
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • Autonomy