基于麦克风阵列的近场声源定位  

Speech source localization in near-field based on microphone arrays

作　　者：李文[1] 夏秀渝[1] 何培宇[1] 李源[1] 

机构地区：[1]四川大学电子信息学院,成都610064

出　　处：《四川大学学报（自然科学版）》2008年第2期307-310,共4页Journal of Sichuan University(Natural Science Edition)

基　　金：国家自然科学基金项目(60472096)

摘　　要：根据近场语音信号的特点,研究使用麦克风阵列的近场声源定位技术.为了克服MUSIC算法本身的局限,对阵列的输入信号进行时-频变换并抽取子带,在此基础上估计声源的数量并对实现对语音信号的定位.最后,根据谱峰的特点,提出了一种快速谱峰搜索算法.仿真结果表明,在有混响的房间中,算法有效的提取了声源的方位.Based on the characteristics of near-field speech signal, This paper researches localization technology using microphone array. To overcome the inherent limitations of MUSIC, we transform the input signals to frequency domain, and then extract sub-band. On this basis, we estimated the number of receive signals and realize localization of speech source. A fast search algorithm is proposed to find the location of speech source. The simulations reso ults show that in a reverberation room, the algorithm effectively finds the position of the speech source.

关 键 词：麦克风阵列 近场语音模型 声源定位 

分 类 号：TP202.7[自动化与计算机技术—检测技术与自动化装置]