Acoustic Detection of Rear Approaching Vehicles for Cyclists

Abstract

Most of the fatal accidents occur due to ignorance and over speeding. Speeding vehicles are more likely to fall prey to unwanted hazards than the slow moving vehicles like bicycles. Notifying the fast rear approaching vehicles is always a problem for a cyclist because looking back frequently is not a good idea. This raises the concern for finding a technical way to help cyclists perceiving the rear approaching vehicles. The increasing safety of cyclists on the suburban roads is important to overcome the hazardous situations. Thus an alerting system needs to be mounted on the bicycle for warning the cyclist. The sensor used to generate the alarm signal should capture the data from the nearby surroundings and compare and analyse the captured data and gives a timely and accurate results in real time. Acoustic detection of vehicles can also works on the principle of Doppler Effect by detecting the Doppler shift. Microphone is chosen as the sensor to record the sound from the rear vehicles and MATLAB is the analysing tool. When a rear vehicle is nearby the cyclist, the microphone records the acoustic signal and this shows peak in the energy. The values are then stored in double precision array. The design of the system includes the generation of audio features that correlates well when a vehicle comes from behind. The point to point analysis is proposed using radar principles as microphone and radar have similar frequency spectrum characteristics in free space.