A comparative study of time delay estimation techniques for road vehicle tracking

Patrick Marmaroli (1) , Xavier Falourd (1) , Hervé Lissek (1)

1 EPFL - Ecole Polytechnique Fédérale de Lausanne

This paper addresses road traffic monitoring using passive acoustic sensing. Recently, the feasibility of the wheelbase length estimation of a road vehicle using one compact microphone array, laid out near the road, has been demonstrated. In essence, the directions of arrival of each incoming sound waves related to the tyre noise are discriminated during the pass-by of the vehicle thanks to a time delay estimation (TDE) technique between signals acquired by distant microphones. The tracking of each DOA is performed thanks to a sequential Monte-Carlo method, also known as particle filtering. This allows the estimation of the trajectory of each wideband moving sound sources; these lasts are identified as axles of the vehicle, the wheelbase length is then recursively estimated. As many TDE techniques are available in the literature, the study investigates their usefulness and their ability to deliver the best tyre noise detection and to feed the tracking algorithm giving kinematics of the axles of the tracked vehicle.

Mots clés: localization; time delay estimation; vehicle tracking; direction of arrival

Domaines: Acoustique [physics.class-ph]

Patrick Marmaroli, Xavier Falourd, Hervé Lissek. A comparative study of time delay estimation techniques for road vehicle tracking. Acoustics 2012, Apr 2012, Nantes, France. ⟨hal-00810981⟩