Helmet Detection for Motorcycle Ignition: A Computer Vision and Android-Based Approach

To enhance safety and combat the motorcycle accident fatality rate, this research introduces a motorbike user authorization system that integrates computer vision technology and an Android application. The system is designed to ensure motorcycle ownership verification while also confirming correct helmet usage by riders. Built on the Raspberry Pi Zero platform, it is capable of processing camera images and communicating with the Android application via Bluetooth. Data, including the speed and accuracy of helmet detection, are analyzed, and assessed using a neural network model and an extensive validation dataset. This discovery underscores the significance of helmet usage in maintaining road safety, given the high incidence of fatal motorcycle accidents associated with rider non-compliance with helmet regulations. With an impressive 90% accuracy rate based on a dataset of 48,000 validation images, this research makes a substantial contribution to mitigating this issue and pioneers a novel approach to bolstering rider safety and compliance with regulations.