Automatic helmet-wearing detection for law enforcement using CCTV cameras

2nd International Conference on Sustainable Development in Civil, Urban and Transportation Engineering

地球科学;运输工程;经济学

Wonghabut, P.^1 ; Kumphong, J.^1 ; Satiennam, T.^1 ; Ung-Arunyawee, R.^1 ; Leelapatra, W.^1

Faculty of Engineering, Khon Kaen University, Khon Kaen

40000, Thailand^1

关键词: C++ language;    Capture images;    Developed applications;    Helmet;    Library use;    Motorcyclist;    Risk behaviour;    Validation results;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/143/1/012063/pdf
DOI  :  10.1088/1755-1315/143/1/012063

来源: IOP

【 摘 要 】

The objective of this research is to develop an application for enforcing helmet wearing using CCTV cameras. The developed application aims to help law enforcement by police, and eventually resulting in changing risk behaviours and consequently reducing the number of accidents and its severity. Conceptually, the application software implemented using C++ language and OpenCV library uses two different angle of view CCTV cameras. Video frames recorded by the wide-angle CCTV camera are used to detect motorcyclists. If any motorcyclist without helmet is found, then the zoomed (narrow-angle) CCTV is activated to capture image of the violating motorcyclist and the motorcycle license plate in real time. Captured images are managed by database implemented using MySQL for ticket issuing. The results show that the developed program is able to detect 81% of motorcyclists on various motorcycle types during daytime and night-time. The validation results reveal that the program achieves 74% accuracy in detecting the motorcyclist without helmet.