Predicting Road Accident Hotspots Using Street-View and Satellite Imagery

Lysaght, Adam Paul (2025) Predicting Road Accident Hotspots Using Street-View and Satellite Imagery. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (5MB) | Preview
Preview	PDF (Configuration Manual) Download (788kB) | Preview

Abstract

Roads facilitate economic and social connections in every day life, yet recurring traffic accidents at certain locations pose safety risks. This paper presents a multi-modal deep learning framework to identify high-risk road segments, or “hotspots”, through the integration and processing of satellite and street-view imagery. A series of machine learning and deep learning models are implemented and evaluated, learning to distinguish visual features associated with historically dangerous road segments. Results demonstrate that combining satellite and street-view data improves classification accuracy by 6% compared to single data source models. The study provides comparative algorithm analysis over a range of machine and deep learning models, offering valuable insights for practitioners aiming to implement targeted interventions.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Anand, Devanshu UNSPECIFIED
Subjects:	T Technology > TE Highway engineering. Roads and pavements Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Motor Industry
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	12 Aug 2026 09:03
Last Modified:	12 Aug 2026 09:03
URI:	https://norma.ncirl.ie/id/eprint/9507

Actions (login required)

View Item