Skip to Main content Skip to Navigation
Theses

Extraction de trajectoires sémantiques à partir de données multi-capteurs : application à des véhicules de secours

Abstract : Positions from GPS sensors are regularly used to analyze the movements of mobile objects. The exclusive use of GPS positions is currently insufficient to detect the behavior of mobile objects because of inaccuracies and loss of information but also because of the semantic richness of these activities. Mobile objects are now equipped with more and more sensors (generic or specific to vehicle types). This thesis introduces a generalizable process of strong semantic enhancement of trajectories obtained from raw data from different sensors. The instantiation of this process provides a tool for semantic trajectory mining, based on the quality analysis of the acquired data, and allowing to identify the activities of vehicles during travel or stops. This process has been implemented for the analysis of the movements of connected ambulances for the Indre et Loire fire department in France. In this context, we used methods based on expert rules defined from the analysis of an ambulance’s working cycle.
Complete list of metadata

https://hal.archives-ouvertes.fr/tel-03365137
Contributor : Frédérick Bisone Connect in order to contact the contributor
Submitted on : Tuesday, October 5, 2021 - 10:45:30 AM
Last modification on : Friday, October 8, 2021 - 3:38:38 AM

File

104249_BISONE_2021_diffusion.p...
Files produced by the author(s)

Identifiers

  • HAL Id : tel-03365137, version 1

Collections

Citation

Frédérick Bisone. Extraction de trajectoires sémantiques à partir de données multi-capteurs : application à des véhicules de secours. Informatique [cs]. Université de tours, 2021. Français. ⟨tel-03365137⟩

Share

Metrics

Record views

44

Files downloads

40