The R*-tree: an efficient and robust access method for points and rectangles
SIGMOD '90 Proceedings of the 1990 ACM SIGMOD international conference on Management of data
The Grid File: An Adaptable, Symmetric Multikey File Structure
ACM Transactions on Database Systems (TODS)
R-trees: a dynamic index structure for spatial searching
SIGMOD '84 Proceedings of the 1984 ACM SIGMOD international conference on Management of data
On map-matching vehicle tracking data
VLDB '05 Proceedings of the 31st international conference on Very large data bases
Map-matching for low-sampling-rate GPS trajectories
Proceedings of the 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems
Hi-index | 0.00 |
Modern mobile technology has enabled the collection of large scale vehicle trajectories using GPS devices. As GPS measurements may come with error, vehicle trajectories are often noisy. A common practice to alleviate this issue is to apply map-matching, i.e., to align vehicle trajectories with the road segments in a digitized road network. This paper presents an efficient solution for map-matching problem that won the SIGSPATIAL CUP 2012. Given a road network, our solution first constructs a gird index on the road segments. For each point p on a vehicle trajectory, we employ the index to identify a candidate set of road segments that are close to p, and then we refine the candidate set to select a segment that matches p with the highest probability. The selection of the best match is based on a metric that takes into account (i) the correlation between consecutive GPS measurements as well as (ii) the directions and shapes of the road segments. Experimental results on real vehicle trajectories and road networks demonstrate the effectiveness and efficiency of the proposed solution.