ANALYSIS, DESIGN AND IMPLEMENTATION OF AN EMBEDDED REALTIME SOUND SOURCE LOCALIZATION SYSTEM BASED ON BEAMFORMING THEORY
Abstract: This project is
intended to analyze, design and implement a realtime sound source localization
system by using a mobile robot as the media. The implementated system uses 2
microphones as the sensors, Arduino Duemilanove microcontroller system with
ATMega328p as the microprocessor, two permanent magnet DC motors as the actuators
for the mobile robot and a servo motor as the actuator to rotate the webcam
directing to the location of the sound source, and a laptop/PC as the
simulation and display media. In order to achieve the objective of finding the
position of a specific sound source, beamforming theory is applied to the
system. Once the location of the sound source is detected and determined, the
choice is either the mobile robot will adjust its position according to the
direction of the sound source or only webcam will rotate in the direction of
the incoming sound simulating the use of this system in a video conference. The
integrated system has been tested and the results show the system could
localize in realtime a sound source placed randomly on a half circle area (0 -
1800) with a radius of 0.3m - 3m, assuming the system is the center point of
the circle. Due to low ADC and processor speed, achievable best angular
resolution is still limited to 25o.
Penulis: Arko Djajadi, Rusman
Rusyadi, Tommy Handoko, Maralo Sinaga, Jürgen Grueneberg
Kode Jurnal: jptkomputerdd090039
