ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-06262023-155019

Tipo di tesi

Tesi di laurea magistrale

URN

etd-06262023-155019

Titolo

Car components detection and recognition for damage detection tasks via Deep Learning

Dipartimento

INGEGNERIA DELL'INFORMAZIONE

Corso di studi

ARTIFICIAL INTELLIGENCE AND DATA ENGINEERING

Parole chiave

car
car components
car damage
cars
components
Deep learning
detection
instance segmentation
recognition
segmentation
vehicle
vehicles

Data inizio appello

21/07/2023

Consultabilità

Non consultabile

Data di rilascio

21/07/2093

Riassunto (Inglese)

Riassunto (Italiano)

The project aimed to find a and train a deep learning model for instance segmentation of car components and for car damages. Specifically, the first part consisted to train several state of the art models with a car components dataset. After a good model was found, it has been improved via genetic algorithms. Then, the model obtained on the first part has been used for the segmentation of car components of damaged cars and has been also retrained with a damaged cars dataset. In this way we can classify correctly damaged areas of a vehicle and recognize which components are involved.

File

Nome file	Dimensione
La tesi non è consultabile. Contatta l’autore