时延估计的声源定位算法及MATLAB实现

Sound source localization algorithm based on time delay estimation with Matlab

概述评价了在机器人、视频会议、大型会议记录中等常用声源定位的时延估计和定位方法,包括广义互相关函数时延估计法、最小方差自适应时延估计法和角度距离定位算法、球形差值定位算法、线性差值法的声源定位方法.利用仿真软件Matlab对各种算法的仿真结果,综合考虑仿真中使用的信号和噪音的互相关性,并结合改进后的相位变换加权函数改进了广义互相关函数时延估计法.结果表明利用改进的广义互相关算法结合球形差值定位技术能够得到更有效、更精确地定位.

Methods of time delay estimation and sound source localization in robots ,video conferences and meeting recordings were reviewed ,including time delay estimation of generalized cross correlation function ,minimum variance of adaptive time delay estimation , position angle distance algorithm , spherical difference localization algorithm and linear interpolation .According to the simulation of vari-ous algorithms with Matlab ,considered the cross correlation feature of input signal and noise in simula-tion ,an improved generalized time delay estimation of cross correlation was presented using the re-formed phase transformation method .The simulation results show that using the improved generalized cross correlation algorithm can obtain more efficient and accurate positioning combined with spherical difference positioning technology .