{"id":{"repo_id":"debrecen","oai_identifier":"oai:dea.lib.unideb.hu:2437/397260"},"canonical_url":"https://search.dev.ndltd.org/etd/debrecen/oai:dea.lib.unideb.hu:2437/397260","repository":{"repo_id":"debrecen","name":"University of Debrecen","base_url":"https://dea.lib.unideb.hu/server/oai/request"},"display":{"title":"Detection of safety vests and face recognition","abstract":"In this paper, we describe a real-time detection system that utilizes a live feed through a camera to identify workers and detect protective safety vests using algorithms based on deep learning. Due to large and complex sites the proposed project will automate the process by providing real-time alerts to safety supervisors by email to ensure corrective action of the violation. The system combines FaceNet, a facial recognition model to identify workers, and YOLOv8n, technology that can analyse real-time video footage to identify personnel and assess safety compliance protocols. The proposed dual functioning system coded with Python and developed in Jupyter Notebook is designed to adapt to a variety of industrial environments. Custom dataset of 15 individual worker images in varying light conditions, different poses, occlusions and with labelled PPE data were made to meet specific algorithmic training requirements. Testing phases have effectively demonstrated positive results of the system, where it recognises faces and detects safety vests with high accuracy of 0.81 for both facial recognition and PPE detection.","abstract_html":"In this paper, we describe a real-time detection system that utilizes a live feed through a camera to identify workers and detect protective safety vests using algorithms based on deep learning. Due to large and complex sites the proposed project will automate the process by providing real-time alerts to safety supervisors by email to ensure corrective action of the violation. The system combines FaceNet, a facial recognition model to identify workers, and YOLOv8n, technology that can analyse real-time video footage to identify personnel and assess safety compliance protocols. The proposed dual functioning system coded with Python and developed in Jupyter Notebook is designed to adapt to a variety of industrial environments. Custom dataset of 15 individual worker images in varying light conditions, different poses, occlusions and with labelled PPE data were made to meet specific algorithmic training requirements. Testing phases have effectively demonstrated positive results of the system, where it recognises faces and detects safety vests with high accuracy of 0.81 for both facial recognition and PPE detection.","abstract_has_math":false,"creators":["Danabek, Magzhan"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":"DE--Műszaki Kar","school":null,"contributors":[],"advisors":["Taleb Abdullah Abdo, Mayar"],"committee_chairs":[],"committee_members":[],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-07-27T19:13:55Z","subjects":["Personal Protective Equipment","YOLOv8n","FaceNet","Real-time Object Detection","Face Recognition;","Deep Learning","Computer Vision","Compliance Monitoring"],"languages":["en"],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/2437/397260","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Taleb Abdullah Abdo, Mayar"]},{"key":"dc:contributor.department","label":"Department","values":["DE--Műszaki Kar"]},{"key":"dc:creator","label":"Author","values":["Danabek, Magzhan"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2025-09-04T15:41:35Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2025-09-04T15:41:35Z"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Personal Protective Equipment","YOLOv8n","FaceNet","Real-time Object Detection","Face Recognition;","Deep Learning","Computer Vision","Compliance Monitoring"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/2437/397260"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["In this paper, we describe a real-time detection system that utilizes a live feed through a camera to identify workers and detect protective safety vests using algorithms based on deep learning. Due to large and complex sites the proposed project will automate the process by providing real-time alerts to safety supervisors by email to ensure corrective action of the violation. The system combines FaceNet, a facial recognition model to identify workers, and YOLOv8n, technology that can analyse real-time video footage to identify personnel and assess safety compliance protocols. The proposed dual functioning system coded with Python and developed in Jupyter Notebook is designed to adapt to a variety of industrial environments. Custom dataset of 15 individual worker images in varying light conditions, different poses, occlusions and with labelled PPE data were made to meet specific algorithmic training requirements. Testing phases have effectively demonstrated positive results of the system, where it recognises faces and detects safety vests with high accuracy of 0.81 for both facial recognition and PPE detection."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["BSc/BA"]},{"key":"dc:title","label":"Title","values":["Detection of safety vests and face recognition"]}]}],"canonical_facts":{"dc:contributor.advisor":["Taleb Abdullah Abdo, Mayar"],"dc:contributor.department":["DE--Műszaki Kar"],"dc:creator":["Danabek, Magzhan"],"dc:date.accessioned":["2025-09-04T15:41:35Z"],"dc:date.available":["2025-09-04T15:41:35Z"],"dc:description.abstract":["In this paper, we describe a real-time detection system that utilizes a live feed through a camera to identify workers and detect protective safety vests using algorithms based on deep learning. Due to large and complex sites the proposed project will automate the process by providing real-time alerts to safety supervisors by email to ensure corrective action of the violation. The system combines FaceNet, a facial recognition model to identify workers, and YOLOv8n, technology that can analyse real-time video footage to identify personnel and assess safety compliance protocols. The proposed dual functioning system coded with Python and developed in Jupyter Notebook is designed to adapt to a variety of industrial environments. Custom dataset of 15 individual worker images in varying light conditions, different poses, occlusions and with labelled PPE data were made to meet specific algorithmic training requirements. Testing phases have effectively demonstrated positive results of the system, where it recognises faces and detects safety vests with high accuracy of 0.81 for both facial recognition and PPE detection."],"dc:description.degree":["BSc/BA"],"dc:identifier.uri":["https://hdl.handle.net/2437/397260"],"dc:language.iso":["en"],"dc:subject":["Personal Protective Equipment","YOLOv8n","FaceNet","Real-time Object Detection","Face Recognition;","Deep Learning","Computer Vision","Compliance Monitoring"],"dc:title":["Detection of safety vests and face recognition"]},"updated_at":"2026-07-27T19:13:55Z"}