Work Place Safety: Investigating and Improving Health and Safety at Construction Sites With Image Recognition Using Deep Learning

Gummaraj Kishore, Mohit (2025) Work Place Safety: Investigating and Improving Health and Safety at Construction Sites With Image Recognition Using Deep Learning. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (760kB) | Preview

Abstract

Construction sites are dynamic and hazardous environments where worker safety depends on the proper use of personal protective equipment (PPE) and correct lifting techniques. This project developed a computer vision system using YOLOv8 to detect PPE compliance and violations in real-time video streams. The dataset included 2,801 annotated images from real-world construction sites, featuring ten object classes that represent compliant and non-compliant safety behaviors. The model was trained using a range of augmentation strategies and tested across multiple epoch configurations to identify the most effective training duration.

The best-performing model was trained for 75 epochs and achieved a mean Average Precision (mAP@50) of 0.49, with a precision of 0.88, and recall of 0.73. The model consistently ran above 50 frames per second when deployed on a Tesla T4 GPU, meeting the requirements for real-time inference. These results show that deep learning based detection systems can support large-scale safety compliance by reducing the need for manual monitoring on construction sites.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Rosales, Charlyn UNSPECIFIED
Uncontrolled Keywords:	Image recognition; PPE detection; posture analysis; YOLOv8; construction safety
Subjects:	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 > Construction Industry H Social Sciences > HD Industries. Land use. Labor > Issues of Labour and Work > Health and Safety at Work.
Divisions:	School of Computing > Master of Science in Artificial Intelligence for Business
Depositing User:	Ciara O'Brien
Date Deposited:	02 Sep 2026 11:49
Last Modified:	02 Sep 2026 11:49
URI:	https://norma.ncirl.ie/id/eprint/9777

Actions (login required)

View Item