Camera-based Intruder Detection and Monitoring of Ship Crew Work Hours

Murad M. M. N., TURGUT B. Ş., Ahmed A., Camliyurt G., Yilmaz Y.

2025 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops, WACVW 2025, Arizona, United States Of America, 28 February - 04 March 2025, pp.1441-1449, (Full Text) identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/wacvw65960.2025.00168
  • City: Arizona
  • Country: United States Of America
  • Page Numbers: pp.1441-1449
  • Keywords: camera ship dataset, crew recognition, intruder detection, ship dataset, work hours estimation
  • Istanbul University Affiliated: Yes

Abstract

Due to technological developments and commercial pressures, the crew size on ships has gradually decreased. As a result of the decrease in crew sizes as well as port and turnaround times, seafarers' work hours have increased and working on ships has become increasingly challenging. The health and performance of fatigued crews deteriorates and the risk of causing accidents increases. The working hour limits of seafarers are specified in international leg-islation, but there is no system that provides evidence to check whether the records of work and rest hours reflect the truth. In this study, a face recognition system is presented using CCTV cameras on ships to monitor the work and rest hours of the crew, and to detect intruders on ships. A new CCTV dataset from a commercial ship is presented to study intruder detection and work hours monitoring.