Urban Objects Classification With an Experimental Acoustic Sensor Network

This paper proposes feature extraction methods for object classification with passive acoustic sensor networks deployed in suburban environments. We analyzed the emitted acoustic signals of three object classes: 1) guns (muzzle blast); 2) vehicles (running piston engine); and 3) pedestrians (several footsteps). Based on the conducted analysis, methods are developed
to extract the features that are related to the physical nature of the objects. In addition, a time-based location method is
developed (based on a pseudo-matched-filter), because the object location is required for one of the feature extraction methods.
As a result, we developed a proof-of-concept system to record and extract discriminative acoustic features. The performance
of the features and the final classification are assessed with real measured data of the three object classes within suburban
environment.