Automated detection of motorcycle helmet use

Abstract

Road traffic collisions are among the top ten causes of death worldwide with more than 1.3 million deaths annually (WHO, 2018). Riders of motorised two- and three-wheelers are more vulnerable to injury and death and make up 28% of global road traffic deaths. In some regions, such as South-East Asia, this number is as high as 43% (WHO, 2018). Correct helmet use reduces the risk of death by 42% and the risk of head injuries by 62% (Liu, Ivers, Blows, Lo, & Norton, 2008). Increasing motorcycle helmet usage to close to 100% by 2030 has been identified as one of the twelve road safety targets by the Global Road Safety Partnership (WHO, 2018). Despite the clear benefits of wearing a helmet, increasing helmet use is challenging especially in low- and middle-income countries (LMICs). A large-scale helmet use media campaign in Thailand over five years showed no benefit (Patummasut, Phewchean, & Sirirattanapa, 2019). While legislating helmet use has shown a clear benefit, there is a disparity between the legislative benefit in high-income countries (HICs) compared to LMICs, with LMICs showing lower use of helmets and less reduction in brain injuries (Lepard, Spagiari, & Park, 2021).