This article proposes a batch-mode algorithm to handle the large databases generated from experimentations using probe vehicles. This algorithm can locate raw Global Positioning System (GPS) positions on a map, but can also be used to correct map-matching errors introduced by real time map-matching algorithms. For each journey, the algorithm globally searches for the closest path to the GPS positions, and so is inspired from the "path to path" algorithm's family. It uses the Multiple Hypothesis Technique (MHT) and relies on an innovative weighting system based on the area between the GPS points and the arcs making up the path. For high performance, the algorithm uses an iterative program and the data is stored in tree form.
Map-matching Algorithm for Large Databases
The journal of navigation ; 68 , 5
2015
Aufsatz (Zeitschrift)
Englisch
Multiresolution descriptor matching algorithm for fast exhaustive search in norm-sorted databases
British Library Online Contents | 2005
|Context-dependent segmentation and matching in image databases
British Library Online Contents | 2004
|Distribution of Large Navigable Databases
British Library Conference Proceedings | 1996
|Promoting safety using retrospective large databases
Kraftfahrwesen | 1998
|Clustering Time-Series Medical Databases Based on the Improved Multiscale Matching
British Library Conference Proceedings | 2005
|