A Sound Source Localization System with Self-calibration Based on Iterative Optimization Algorithm

A seven-microphone array with three microphones in each axis is designed. The array is developed to localize acoustic source of near and far field in three-dimensional space. Moreover, it can achieve self-calibration. The algorithm is composed of two sections. In the first section, the traditional generalized cross correlation is used to obtain the time-delay among each microphone. In the second section, we propose a novel resolving algorithm based on the iterative optimization. Accurate sound source localization can be realized by approaching the target with the numerical solution constantly. Experimental results demonstrate a good performance in localization. It should be emphasized that the space of RAM and ROM occupied by algorithm is small and the speed of computation is fast. Last but not least, traditional approximate formulas are based on the localization of far-field, but the iterative optimization algorithm can be applied accurately for both far-field and near-field.