.png){kind=icon}
.png){kind=icon}
Return
Efficient path decoding from high sampling trace data using TSPS
.png){kind=icon}
PRE
AI
DOI：10.1080/10095020.2025.2548370.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
With reduced storage costs and increased network bandwidth, map matching (MM) or path decoding from high sampling trace data for all private and shared vehicles, bicycles, etc., is anticipated in the near future. To save the storage space, traditional MM methods are often designed in a step-by-step manner for low sampling trace with a time interval above 1 min. But the step-wise matching logic is naturally inefficient for high sampling trace. We propose integrating all path developing work into only one trace-oriented shortest path search (TSPS). Five existing MM algorithms with different speedup strategies are used to benchmark the performance of TSPS. The experiment results conducted on two trajectory datasets validated that the proposed algorithm has an outstanding working efficiency by up to four orders of magnitude without loss of accuracy.
Keywords:
Urban computing
map matching
efficient path decoding
high sampling trace data
trace-oriented shortest path search
Journal
G
IF:
5.5
Papers:
837
Citations:
2.4K