Skip to Main content Skip to Navigation
Theses

Navigation des personnes aux moyens des technologies des smartphones et des données d’environnements cartographiés

Abstract : Smartphone navigation using the low-cost embedded sensors in off the shelf smartphones can provide a continuous solution in GNSS-denied environments. The most widely adopted approach is Pedestrian Dead Reckoning (PDR) that uses acceleration and angular velocity to estimate the user’s position. Yet, consumer grade sensors deliver noisy measurements that may result into a drift in the estimated trajectory. One major challenge is to estimate accurately step length information since it depends on physiological features that are specific to each user. In addition, angular biases are more likely to be introduced in the orientation estimation process with handheld devices. This is mainly due to the high degree of freedom of hand motion. In the context of a national project called HAPPYHAND, the main goal of this work is to exploit map information as far as possible in order to mitigate the previous inherent limitations to the PDR approach. First, a topological network extracted from the map is proposed in order to correct the angular errors and calibrate the step length model. Second, context awareness is adopted in order to provide regular and frequent position updates thanks to a point of interest online detection scheme.
Complete list of metadatas

Cited literature [128 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02157807
Contributor : Abes Star :  Contact
Submitted on : Friday, May 3, 2019 - 5:17:35 PM
Last modification on : Friday, July 17, 2020 - 5:09:11 PM

File

F_TAIA_ALAOUI.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02157807, version 2

Collections

Citation

Fadoua Taia Alaoui. Navigation des personnes aux moyens des technologies des smartphones et des données d’environnements cartographiés. Traitement du signal et de l'image [eess.SP]. École centrale de Nantes, 2018. Français. ⟨NNT : 2018ECDN0051⟩. ⟨tel-02157807v2⟩

Share

Metrics

Record views

252

Files downloads

617