摘要人类最常用、最重要、最方便和最有效的交换信息的形式是通过语音传递信息，因此语音信号也是人们思想沟通和交流感情的最主要途径。现在，人类已开始进入信息化时代，用现代化手段研究语音处理技术，可以使人们更加有效地产生、传递、存储、获取和应用语音信息，这对于促进当今社会的发展具有十分重要的意义。64614

语音时域分析方法直接对语音信号的时域波形进行分析，是最简单直观的方法，提取语音的短时能量和短时过零率等特征参数。本文将对这几种时域参数和其用途进行详尽的介绍，用C++语言设计了一个绘制语音波形、提取短时能量和短时过零率，并用短时能量和短时过零率进行语音信号端点检测的程序，实验结果表明综合应用短时能量和短时过零率能较好地检测静音段与语音段。

毕业论文关键词  语音信号 短时能量 短时过零率 端点检测

毕业设计说明书（论文）外文摘要

Title      Speech time domain characteristics analysis   

Abstract    Transmission of information through Speech signal is the most important, effective,commonly used and most convenient form of exchange information, and therefore speech signal is the main way to constitute of a clear thought and emotional communication. Nowadays, mankind has begun to enter the information age, using modern means to study speech processing technology, can make people more effectively to produce, transport, storage, and application of voice information, it takes great significance to promote the development of the society.

Speech time domain analysis method is the simplest and most intuitive method, it analyses the speech signal time domain waveform directly, and the parameters mainly include the voice of short-time energy and short-time zero crossing rate, etc. In this paper, these time domain parameters and their use will be introduced in detail and design a program which can draw speech waveform,extract short-time energy and short-time zero crossing ratio and detect the speech signal endpoints based on short-time energy and short-time zero crossing ratio.Experimental results show that the integrated application of short-time energy and short-time zero crossing ratio can well detect quiet voice segment and voice segment.

Keywords: Speech signal ,short-time energy, short-time zero crossing rate, endpoint detection

1 绪论 1

1.1 语音信号处理研究现状 1

1.2本文研究内容及意义 2

1.3本文组织结构 2

2 语音信号处理的基础知识 3

2.1 语音信号的波形特性 3

2.2 语音的分类 4

2.3 汉语语音的基本特征 5

2.3.1 汉语语音的特点 5

2.3.2 声母和韵母 5

2.3.3 元音和辅音 6

2.3.4 汉语的四声 7

3 语音信号的时域特征分析 7

3.1 语音信号的预处理 8

3.1.1 语音信号的预加重处理 8

3.1.2 语音信号的加窗处理 8

3.2 短时平均能量分析 10

3.3 短时平均过零率分析 11

3.4 基于能量和过零率的语音端点检测