Abstract

This paper proposes a map building algorithm for navigation task of autonomous vehicle. A local 2D grid map is designed to convey three types of information (free space, obstacle and unknown). A refined RBPF-SLAM algorithm is established to build a large scale, accurate navigation map. Moving obstacle information is excluded in the final map, which provides better performance in vehicle's online localization. Two sets of experiments are carried out in two driving conditions, with one in urban traffic environment and the other in unstructured environment. The mapping and localization result testify the effectiveness of our algorithm.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call