Simple Sound Source Detection Using Main-Lobe Model of Microphone Array

摘要

The paper proposes simple sound source detection method using the main-lobe model of a microphone array. The method selects main-lobe part of particular sound source from spatial spectrum obtained from Beam Forming and localizes multiple sound sources without using reflected or interfered components. We implement the method to out 32ch microphone array, and evaluate the performance in different conditions. The efficiency in reverberant environment or on mobile robot is also shown.