Microphone array based sound source detection and localization with beamforming, gcc-phat, and triangulation using arduino and MATLAB
Microphone arrays are becoming popular for sound source localization (SSL). Keeping real-time applications in mind, such as search and rescue missions, military operations for gunshot detection, scream based localization, etc., it will help us understand how important sound localization is for people’s safety. This research adopts a simple and powerful, Arduino based MATLAB sound source detection and localization system, with a reduction in localization error. Sound source localization rely mostly on an estimation of time of arrival (TOA), direction of arrival (DOA) and time difference of arrival (TDOA) methods at microphone sound sensors. The signals are then analyzed by cross-correlation functions and spatial gradient approach or time-delay based methods are applied, which will be described in our methodology. A tap-based sound source is used to detect and analyze the sound locations data. Testing of the whole system is embedded with MATLAB to increase the performance and locate the source point on MATLABGUI. The data form microphone sound sensors are collected by Arduino Uno and analyzed in MATLAB. The number of actual tap locations to that of estimated tap locations using GCC-PHAT and Sound triangulation methods are collected. The results are compared between software and hardware data tap localizations. In this approach of microphone array-based sound source detection and localization system, the accuracy is checked by calculating the localization errors, which is a viable path in this research work.
Sound source detection, Localization
Kadukuntla, T. (2019). <i>Microphone array based sound source detection and localization with beamforming, gcc-phat, and triangulation using arduino and MATLAB</i> (Unpublished thesis). Texas State University, San Marcos, Texas.