Open AccessHelmet wear detection based on neural network algorithm
Author(s) -
Ruiyun Cao,
Hongxiang Li,
Banglie Yang,
Ao Feng,
Jie Yang,
Jiong Mu
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1650/3/032190
Subject(s) - task (project management) , computer science , artificial neural network , variety (cybernetics) , work (physics) , face (sociological concept) , artificial intelligence , machine learning , real time computing , engineering , systems engineering , mechanical engineering , social science , sociology
Wearing safety helmet correctly is an effective means to reduce the accident rate and ensure the safety of construction personnel. But in the complex work site, the helmet detection task will face a variety of challenges. This paper uses YOLOV4’s deep neural network architecture and makes fine-tuning and improvement according to the characteristics and difficulties of this task, proposes a new helmet detection system, and achieves 95.1% accuracy. The application can timely and accurately detect whether the personnel wear the helmet correctly, which is of great significance for ensuring construction safety in practical work.