An approach for sound sources localization and characterization using array of microphones

In this paper, sound sources are identified and localized by using the time domain beamforming and geometrical approach. The location of the dominant sound source is identified by the conventional delay-and-sum beamforming approach. Subsequently, the acoustics signal of such a dominant sound source can be characterized by multiplying corresponding time delay functions from the source to each sensor. In order to obtain a more accurate acoustic signal of the source, an approach of elimination of the sidelobe which is contributed by other sound sources is developed. Ultimately, the receiving signal can be reconstructed such that the feature of the acoustics signals of each sound source can be obtained. Multiple loops of conventional beamforming require to be performed in order to separate and localize all the main sound sources in the sound field one by one. The approach has been validated by experiment. In order to shorten the computation time due to the searching of the location of the sound sources accurately by beamforming approach, an increasing-spatial-resolution scanning strategy near the sound field of the sound source is also adopted in the current study. The proposed approach gives promising result of the sound source localization and separation in free field condition.