Master's Thesis

Comparative analysis of map matching approaches

Gonçalo Costa Barbas — 2022

Key information

Authors:

Gonçalo Costa Barbas (Gonçalo Costa Barbas)

Supervisors:

Francisco António Chaves Saraiva de Melo (Francisco António Chaves Saraiva de Melo); António João Ferreira Reis

Published in

November 10, 2022

Abstract

The task of map matching stands for the location of road segments that correspond to some GNSS measurements, to find the route traversed by a vehicle. This is growing relevance for Intelligent Transportation Systems, GNSS-oriented navigation systems, and other real-life tools. Thus, several approaches have been proposed towards achieving progressively more accurate results, encouraging the urge to a comparative analysis among them. This work conducts a clear exposition of the main classes of approaches and a subsequent quality evaluation, as well as a definition of good performance metrics. The results showed that geometrical approaches (that make use of geometry of the roads), despite simple and fast in running time, are very sensitive to the road structure, often leading to unfeasible paths. The topological approaches use information concerning the connectivity between roads, considering the contiguity of the journeys. These models brought some improvement to the previous but tend to fail for higher sampling periods. HMM are a known structure for modeling sequential processes, to predict the most likely sequence of states, introducing a probabilistic approach to map matching. This model achieved better results than the earlier, but still reaches a significant mismatch percentage for noisy data. HMM can use Baum-Welch procedure when ground-truth data is not provided, despite the costs in terms of efficiency. Finally, the attentional sequence-to-sequence model handles data with poorer quality and can be trained with generated data, by using data augmentation methods.

Publication details

Authors in the community:

Supervisors of this institution:

Fields of Science and Technology (FOS)

electrical-engineering-electronic-engineering-information-engineering - Electrical engineering, electronic engineering, information engineering

Publication language (ISO code)

eng - English

Rights type:

Embargo lifted

Date available:

October 7, 2023

Institution name

Instituto Superior Técnico