Acoustical Classification of the Urban Road Traffic with Large Arrays of Microphones

Abstract

This work is part of a study dealing with city-dwellers' quality of life. Noise is known to be an important factor influencing the quality of life. In order to diagnose it properly, we propose a noise monitoring system of urban areas. It is based on the use of large microphone arrays in order to extract the radiated sound field from each passing-by vehicle in typical urban scenes. A machine learning algorithm is trained so as to classify these extracted signals in clusters combining both the vehicle type and the driving conditions. This system makes it possible to monitor the evolution of the noise levels for each cluster. The proposed system was first tested on passing-by isolated vehicles measurements and then implemented in a real street in Paris (France).