短时能量和短时过零率的分析工作是语音信号的时域分析中最基本的研究方法,同时它们也有着相当广泛的应用,这一点,在语音信号的检测方面尤其明显。早期的做法是讲两者分开检测,相互独立使用。生产实践结果表明这样的做法很容易出现差错,不利于有效的识别语音信号。本文将两种时域分析的方法结合起来,利用MFC框架和VS平台进行程序设计将短时能量和短时过零率两个时域信号的分析结合在一起,通过双门限法处理信号,对端点检测方法进行了研究。实验结果表明这样的分析方法比用一个特征参数进行端点检测更加精确。81873
毕业论文关键词 短时能量 短时过零率 双门限法 端点检测
毕业设计说明书(论文)外文摘要
Title Short-term energy speech signal And short-term zero rate Characteristics
Abstract Short-term energy analysis and short-term zero rate is a time-domain speech signal analysis methods in basic research, but they also have a wide range of applications, which, in speech signal detection is particularly evident。 Early detection approach is about two separate, independent use。 Practice results show that this approach is prone to error, is not conducive to effective recognize speech signal。 In this paper two methods combine time-domain analysis, the use of MFC framework program and VS platform to build model, short-term energy analysis and short-term zero rate of two time domain signals to blend together, by double threshold method in signal processing, endpoint detection methods were studied。 This analysis method is more accurate than using a signal to carry out endpoint detection。
Keywords Short-term energy speech signal short-term zero rate Double threshold method Endpoint detection
目录
1 绪论 1
1。1 课题设计背景 1
1。2 课题设计目的和任务 2
2 语音信号的时间域特征分析 3
2。1 语音时域信号的特点 3
2。2 短时能量 3
2。3 短时过零率 4
2。4 WMV5
3 开发工具及开发技术介绍 6
3。1 C++简介 6
3。2 MFC框架简介 6
3。3 Visual Studio 2010简介8
4 语音信号时间域特征分析程序设计 9
4。1 需求分析 10
4。2 程序设计思路 10
4。3 界面设计 17
4。4 语音信号的特征分析及端点检测 20
4。4。1 短时能量计算