In-Vehicle Safe Driving Aid and Inter-Vehicle Interaction Technology

Abstract

Management of road safety and traffic efficiency is the responsibility of all road users. However, safe driving is no longer possible for many, especially when travelling on either long-distance trips or under scotopic conditions when the ambient illumination is limited. In this project we developed a safe driving intelligent system with near real-time inter-vehicle communication capabilities. This system allowed the acquisition and sharing of images of road conditions and their potential risk via Google Cloud service with other road users. The quantitative analysis of images acquired at different times of the day revealed a high correlation between the image Signal to Noise Ratio (SNR) and illumination on the detector. Our results recorded the highest mean SNR value of 25.1 ± 0.5 dB at noon time before the value gradually declined as the day approached evening. Meanwhile, an investigation into the quality of images taken from a vehicle travelling at a range of speeds revealed a relatively high inconsistency in the mean SNR value of 9.1 dB. We attribute this problem to the low frame rate of the employed imaging system. The fast system response time of 5.6 ± 1.65 s that allows rapid image acquisition and immediate distribution makes it a feasible and useful driving aid for a safer driving experience, and to increase the safety of the community.