基于MATLAB的语音信号识别及矢量模式匹配  

Identifying Sound Signal and Mode-matching of Vectors Based on MATLAB

作　　者：承江红[1] 邰超[2] 肖登明[2] 

机构地区：[1]广州大学物理与电子工程学院,广东省广州市510006 [2]上海交通大学电子信息与电气工程学院,上海市200030

出　　处：《微计算机信息》2012年第10期443-445,共3页Control & Automation

摘　　要：语音识别技术一直受到各国科学家的广泛关注,如今,随着语音识别技术研究的突破,其对计算机发展和社会生活的重要性日益凸现出来。本文在MATLAB平台上,将语音文件输入,经过时域处理、频域处理、线性预测法提取特征向量,最后进行矢量量化模式匹配。重点是对语音信号进行线性预测分析和矢量量化的模式匹配。The technology of identifying sound signal has always been drawing much attention from the scientists worldwide, nowadays, with the breakthrough of the technology of identifying sound signal, its importance for the development of computers and social life is shown especially out day by day. This paper, based on MATLAB, inputs the audio file, extracts the feature vector by time domain processing, frequency domain processing and linear prediction and matchs the mode of vector quantization at last. The key point is linear prediction analysis and mode-matching of vector quantization of voice signal.

关 键 词：MATLAB 语音识别 时域分析 LPC VQ技术 

分 类 号：TP391.42[自动化与计算机技术—计算机应用技术]