This thesis deals with the autonomous driving topic and it is a complete guide on how to design and implement an autonomous driving system on a scale vehicle. The project starts from the requirements analysis to the final product consisting of a model that can complete a navigation task, reaching an objective location in an unknown environment, avoiding obstacles and recognizing road signals. The project is considered to be a base system that can be used as a starting point for future developments; it is based on the Robot Operating System [ROS] framework and the used hardware consists of an RC vehicle model, a Raspberry Pi 4, an IMU, a lidar and a camera as sensors. To design all the tasks involved in an autonomous navigation system the used algorithms are Google Cartographer as Simultaneous Localization and Mapping [SLAM] method, Dijkstra and Timed Elastic Band [TEB] as global and local planners, and finally YOLOv8 as detection model to recognize road signals.
Questa tesi tratta l’argomento della guida autonoma ed è una guida completa su come progettare ed implementare un sistema di guida autonoma su di un veicolo in scala. Il progetto parte dall’analisi dei requisiti, per arrivare al modello finale che è in grado di completare una navigazione autonoma, raggiungendo la posizione di un obiettivo in un ambiente sconosciuto, evitando ostacoli e riconoscendo i cartelli stradali. Il progetto è inteso come un sistema base da usare come punto di partenza per futuri sviluppi; è basato sul framework Robot Operating System [ROS] e l’hardware usato consiste in un modello di veicolo radiocomandato, un Raspberry Pi 4, una IMU, un lidar e una camera. Per progettare tutti i sottoinsiemi che compongono un sistema di guida autonoma, gli algoritmi usati sono Google Cartographer come metodo di Simultaneous Localization and Mapping [SLAM], Dijkstra e Timed Elastic Band [TEB] come global e local planners, infine YOLOv8 come modello di percezione per riconoscere i cartelli stradali.
Autonomous driving and traffic sign recognition: build of a scale vehicle
MILAN, FEDERICO
2023/2024
Abstract
This thesis deals with the autonomous driving topic and it is a complete guide on how to design and implement an autonomous driving system on a scale vehicle. The project starts from the requirements analysis to the final product consisting of a model that can complete a navigation task, reaching an objective location in an unknown environment, avoiding obstacles and recognizing road signals. The project is considered to be a base system that can be used as a starting point for future developments; it is based on the Robot Operating System [ROS] framework and the used hardware consists of an RC vehicle model, a Raspberry Pi 4, an IMU, a lidar and a camera as sensors. To design all the tasks involved in an autonomous navigation system the used algorithms are Google Cartographer as Simultaneous Localization and Mapping [SLAM] method, Dijkstra and Timed Elastic Band [TEB] as global and local planners, and finally YOLOv8 as detection model to recognize road signals.File | Dimensione | Formato | |
---|---|---|---|
2024_10_Milan.pdf
accessibile in internet per tutti a partire dal 27/08/2025
Descrizione: Tesina magistrale - Federico Milan
Dimensione
4.55 MB
Formato
Adobe PDF
|
4.55 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/224954