Design and Implementation of an Intelligent Motorcycle Helmet for Large Vehicle Approach Intimation

Abstract

This paper proposes an intelligent motorcycle helmet that includes two miniature infrared transceivers, a miniature image sensor (camera), an embedded computation processing (ECP) module, a battery charging module, a microphone, and earphones. Two large truck/bus vehicle registration plate image recognition modes (day and night) are designed and implemented for the purpose of large truck/bus approach intimation. In total, 600 images of rear-approaching trucks/buses on an actual road from 10 motorcyclists during day and night conditions are used to test the accuracy of the vehicle registration plate recognition. The experimental results show that the intelligent helmet proposed can successfully recognize an image of the vehicle registration plate of a large truck/bus in Taiwan. The accuracy of the vehicle registration plate recognition is up to 75% during the day and 70% at night. As a result, the intelligent motorcycle helmet proposed can detect the vehicle registration plate of an approaching large truck/bus in real time within a distance of 5 m.