Traffic monitoring with ad-hoc microphone array

Abstract

In this paper, we propose an easy and convenient method for traffic monitoring based on acoustic sensing with vehicle sound recorded by an ad-hoc microphone array. Since signals recorded by an ad-hoc microphone array are asynchronous, we perform channel synchronization by compensating for the difference between the start and the end of the recording and the sampling frequency mismatch. To monitor traffic, we estimate the number of the vehicles by employing the peak detection of the power envelopes, and classify the traffic lane from the difference between the propagation times of the microphones. We also demonstrate the effectiveness of our proposed method using the results of an experiment in which we estimated the number of vehicles and classified the lane in which the vehicles were traveling, according to F-measure.