In this paper we address the problem of detecting road intersections. We present two approaches to solve the problem of intersection detection in an unstructured outdoor setting. The first is a natural extension of the popular VFH* obstacle avoidance algorithm. It detects intersections and tracks, over a period of time, the angles at which gaps in the robot's certainty grid (CG) are first observed. The second approach uses techniques from image processing and computational geometry on the certainty grid image, to extract a skeleton of the navigable region, thus providing the intersections. We show experimental results portraying intersection detection due to both methods and show the results. On the whole, we found that the robot was able to detect all possible intersections.
Outdoor intersection detection for autonomous exploration
2012-09-01
1362825 byte
Conference paper
Electronic Resource
English
Autonomous Intersection Management
NTIS | 2009
|INTERSECTION POSE DETECTION IN AUTONOMOUS MACHINE APPLICATIONS
European Patent Office | 2020
|INTERSECTION DETECTION AND CLASSIFICATION IN AUTONOMOUS MACHINE APPLICATIONS
European Patent Office | 2020
|INTERSECTION DETECTION AND CLASSIFICATION IN AUTONOMOUS MACHINE APPLICATIONS
European Patent Office | 2024
|Intersection detection and classification in autonomous machine applications
European Patent Office | 2024
|