Acoustic Detector of Road Vehicles Based on Sound Intensity

Abstract

A method of detecting and counting road vehicles using an acoustic sensor placed by the road is presented. The sensor measures sound intensity in two directions: parallel and perpendicular to the road. The sound intensity analysis performs acoustic event detection. A normalized position of the sound source is tracked and used to determine if the detected event is related to a moving vehicle and to establish the direction of movement. The algorithm was tested on a continuous 24-h recording made in real-world conditions. The overall results were: recall 0.95, precision 0.95, F-score 0.95. In the analysis of one-hour slots, the worst results obtained in dense traffic were: recall 0.9, precision 0.93, F-score 0.91. The proposed method is intended for application in a network of traffic monitoring sensors, such as a smart city system. Its advantages include using a small, low cost and passive sensor, low algorithm complexity, and satisfactory detection accuracy.

Traffic Noise Assessment Using Intelligent Acoustic Sensors (Traffic Ear) and Vehicle Telematics Data

“…For example, Ishida et al and Uchino et al used a method that relied on the time difference of arrival (TDOA) of sound signals on two microphones on a sidewalk to detect the vehicles [ 32 , 33 ]. Several other researchers demonstrated the use of a new method based on the analysis of sound intensity signals measured in a two-dimensional space to determine the average speed of the vehicle and the direction of traffic flow [ 34 , 35 , 36 ]. Other approaches to acoustic vehicle detection have been constructed on machine learning-based techniques, as discussed by Gatto and Forster [ 37 ].…”

“…Essentially, the total intensity of the moving object was measured first, and the source position was then determined using a Kalman filter in a method that was previously proposed by Szwoch and Kotus [ 36 ]. The object was then considered by the sensor if its total intensity was found to be higher than the usual noise of road vehicles.…”

“…The object was then considered by the sensor if its total intensity was found to be higher than the usual noise of road vehicles. Together, with our real-world measurements, it has been shown that the sound wave energy (total) emitted by moving road vehicles is concentrated in the mid-frequency range of 50-200 dB [36]. In order to avoid false-positive recordings, only the objects located in the front area of the sensor with a total length of 85 m were considered.…”

“…of traffic flow [34][35][36]. Other approaches to acoustic vehicle detection have been constructed on machine learning-based techniques, as discussed by Gatto and Forster [37].…”

“…Vehicle miles here refer to the total distance travelled by all vehicles during the year [45], while road length is given as the actual length of the roads in the area. Essentially, the total intensity of the moving object was measured first, and the source position was then determined using a Kalman filter in a method that was previously proposed by Szwoch and Kotus [36]. The object was then considered by the sensor if its total intensity was found to be higher than the usual noise of road vehicles.…”