摘要随着信息技术和多媒体技术的发展，实现人机的语音交互已成我们人人日益迫切的需求。语音信号处理技术不断发展，语音信号中包含着大量的有用信息的同时也包含着大量的冗余信息，正确的选择语音信号中的特征参数来表征语音信号一直是该技术研究的主要内容之一。正确高效的语音信号流将加速通信技术的发展，同时将加速人机交互的实时性与高效性，这必将是未来语音信号处理技术发展的方向。现代语音信号特征参数的研究主要是线性预测系数（LPCC）和Mel倒谱系数（MFCC)或是基于两者的综合。语音信号特征参数提取一直是语音信号处理研究中主要的研究方向。63606

本文目的是对语音信号特征参数进行提取，并验证参数的有效性。本文对语音信号进行了时域和频域的理论分析。同时对语音信号处理中的多个过程进行了理论介绍与分析，对语音信号的时域频域的各项参数进行了简单分析与Matlab仿真。最终通过Matlab仿真实现了对语音信号的LPCC和MFCC参数提取。

毕业论文关键词：语音信号, 特征参数提取, LPCC ，MFCC,Matlab仿真等

毕业设计说明书（论文）外文摘要

Title Research  about  characteristic  parameters of   the   speech    recognition                    

Abstract With the development of information technology and multimedia technology, the realization of human voice interaction has become increasingly urgent demand for us . Speech signal processing technology continues to evolve. The speech signal contains not only a lot of useful information, but also a lot of redundant information.So,in the speech analysis,  the correct speech signal characteristic parameters has been one of the main research . Correct and efficient voice signal flow will accelerate the development of communication technology, and also will accelerate the real-time human-computer interaction and efficiency, which will be the direction of future voice signal processing technology approaching . Modern speech signal characteristic parameter study is based on linear prediction coefficients (LPCC) and Mel cepstral coefficients (MFCC) or even the combination of the two. The characteristic parameter extraction of speech signal  is the main  research direction of speech signal processing,as always.

Purpose of this paper is to extract characteristic parameters of the speech signal, and verify the validity of parameter. This paper first introduced the speech signal theory in time domain and frequency domain then presented the voice signal processing of multiple processes theory.After analyzed the feature of speech signals in detail, we used Matlab software to simulate the algorithms.Finally, the LPCC and MFCC parameter of the speech signal are extracted.

Key words: speech signal, characteristic parameters extraction, LPCC,    MFCC, Matlab simulation

目次

第一章 绪 论 1

1.1 语音信号处理技术简介 1

1.2 国内外研究现状及论文选题意义 3

1.3 论文主要研究内容及结构安排 3

第二章 语音信号分析 4

2.1 概述 4

2.2 语音数字化和预处理 4

2.2.1 数字化 4

2.2.2 预处理 5

2.2.3 窗函数的选择 5

2.3 语音信号的时域分析 7

2.3.1 信号能量、平均幅度分析