摘要
基于时延估计的声源定位算法是先求得时延,再根据时延来计算出声源的位置。因此时延估计的准确性决定了声源定位的准确性,而时延估计的精度会受到室内混响和噪声的影响,因此本文提出了一种改进后的PHAT加权算法。通过传感器矩阵模型采集声源的音频信号传到上位机并在Matlab进行算法仿真。结果表明改进后的PHAT加权算法结合牛顿迭代定位算法能够比较准确的对声源进行定位。
Based on the time-delay estimation of the sound source localization algorithm is to find the delay, and then calculate thesound source according to the delay position. Therefore, the accuracy of the delay estimation determines the accuracy of the soundsource localization, and the accuracy of the delay estimation will be affected by the indoor reverberation and noise. Therefore, thispaper presents an improved PHAT weighting algorithm. Through the sensor matrix model, the audio signal of the sound source iscollected and transmitted to the host computer and the algorithm is simulated in Matlab. The results show that the improved PHATweighting algorithm combined with Newton iterative positioning algorithm can accurately estimate the sound source Positioning.
作者
李伟
LI Wei （School of Electrical and Information Engineering, Wuhan Institute of Technology, Wuhan 430074 ,China）
出处
《电脑知识与技术》
2018年第3期220-222,共3页
Computer Knowledge and Technology
关键词
时延估计
广义互相关
声源定位
牛顿迭代
Delay estimation
Generalized mutual correlation
Source localization
Newton iteration
