Logo image
運用多個聯網麥克風進行室內環境語音與音樂之增強:波束成形方法開發與評估
Thesis

運用多個聯網麥克風進行室內環境語音與音樂之增強:波束成形方法開發與評估

陶柏戎
Masters, 國立清華大學, 電機工程學系
2015

Abstract

聲學 增強 Acoustic Enhancement
Acoustic signals play an important role in both multimedia and communication,and acoustic signal enhancement using beamforming is still a challenging issue, especially for communication equipments. Two main methods are frequently used in acoustic beamforming : 1. Delay and Sum 2. Minimum Variance Distortionless Response (MVDR). In this research, we intend to construct an acoustic beamforming system to implement the two methods in the indoor environment by using multiple microphones and a low-cost single board computer, Raspberry Pi. For the estimation of time difference of arrival (TDOA), fractional delay is calculated to increase the resolution in time. In order to improve the acoustic enhancement, we propose a method that determines weights to make the power of noise leveled across different channels. We assume the noise is uncorrelated across channels, so the signalto-noise ratio can be improved simply by summation if signals are accurately aligned in time. The experiment result shows that the proposed method of determining the weights successfully enhances the signal in the indoor environmentand increases signal to noise ratio (SNR) in comparison with the conventional Delay and Sum method .

Metrics

1 Record Views

Details

Logo image