Sound Source Localization Based On Fpga

Sound localization has broad prospects, and has been concerned by many researchers. There is a wide range of applications from a single source location to the combined audio and video information, and its implementations and algorithm have become increasingly diverse.This dissertation details sound location system based on FPGA development platform. Using this method can achieve module design, and it is convenient for system migration and to combine with other fact information systems. In this dissertation, a variety of sound source location algorithms are introduced. The system adopts the algorithm based on the time difference of arriva (TDOA). Compared with the existing microphone array sound source location methods, this algorithm, commonly used in sound source location system, needs less computation but with higher positioning accuracy and is easy to realize real-time. The positioning system is implemented on the development board Virtex4, after capturing sound source and enlarging sound signals using the front-end circuit of microphone, it experiences A/D conversion by AIC32B, and then the digital sound signal is sent to FPGA and processed by it. The dissertation compares different development programs based on FPGA, and chooses the traditional HDL language development program, taking advantage of the superiority that FPGA can be deployed, is implemented on the hardware. And the method simplifies the system structure and raises the processing speed.Through testing the sound source location system, the system reaches indicators of real-time, reliability and stability both in hardware and software, and it has good portability and versatility in function.