The purpose of the map-matching module of an in-vehicle navigation system is to locate the position of the vehicle relative to the map data that is referenced by the system. Map-matching allows the display module to accurately depict to the driver the position of the vehicle on the map, facilitates operations such as route calculation, supports ADAS applications such as adaptive cruise control (ACC) and is required for the ultimate goal of intelligent vehicle research, viz., autonomous driving. The basic approach is to compare the historical vehicular path to possible candidate paths on the map, where candidate paths are the paths on the map which lie within the general vicinity of the vehicle at any instant in time. The candidate map path which best matches the vehicle path is the most likely path on which the vehicle is located. This paper presents two novel metrics, the spatial angle variation metric and the iterative multiresolution trend metric, which provide new methods for two dimensional and three-dimensional map-matching respectively. The methods described here are effective while being both intuitive and computationally non-intensive and provide new insight into the problem of map-matching.
Novel metrics for map-matching in in-vehicle navigation systems
2003
8 Seiten, 10 Quellen
Aufsatz (Konferenz)
Englisch
Novel Metrics for Map-Matching in In-Vehicle Navigation Systems
British Library Conference Proceedings | 2003
|Fast stereo matching for autonomous vehicle navigation
IEEE | 2005
|Depth Image Matching for Underwater Vehicle Navigation
British Library Conference Proceedings | 1999
|Performance metrics for robot navigation
IEEE | 2007
|