基于广义互相关时延估计算法声源定位的研究
首发时间:2010-08-25
摘要:在现有的传声器阵列声源定位方法中,基于声达时间差( TDOA) 估计定位法计算量较小,定位精度较高,同时也易于实现实时系统,是目前声源定位法中常用的方法。采用该方法最重要的就是进行时间延迟估计( TDE) ,其精确性直接影响到定位的准确与否。为了精确的得到时延估计,本文首先介绍基于时延估计的声源定位算法以及其实验研究,进而对广义互相关延时估计算法中几种加权函数的性能进行了分析,给出了部分加权的仿真结果,分析了每种算法中存在的优缺点,通过比较,PHAT加权在较高的信噪比时,具有波动小、峰值尖锐的特性,在降低信噪比时,又有较强的抗干扰特性,是基于广义互相关时延估计算法声源定位的最佳选择。?????
For information in English, please click here
Time-delay estimation based on generalized cross-correlation algorithm for acoustic source localization
Abstract:In the existing Microphone array sound source localization methods, TDOA the commonly used method is simple, accuracy and easy to implement real-time systems. In this method the most important thing is Time Delay Estimation (TDE) that directly affects the accuracy of positioning. In order to obtain accurate TDE, we introduce the sound source localization algorithm based on TDE,and then analyze the performances of several weighted functions in the generalized cross correlation delay estimation algorithm and give the simulation results and the advantages and shortcomings of each function. By comparison, PHAT weighting is characterised by small fluctuationsa,low sharp peak in the high SNR and strong anti-jamming feature in the lower SNR.In the generalized cross correlation delay estimation algorithm PHAT is the best choice for sound source localization.
Keywords: AE Time-delay estimation acoustic source localization;GCC
基金:
论文图表:
引用
No.4382250203363128****
同行评议
共计0人参与
勘误表
基于广义互相关时延估计算法声源定位的研究
评论
全部评论0/1000