基于基音周期模型的语音识别毕业设计

《基于基音周期模型的语音识别毕业设计》由会员分享，可在线阅读，更多相关《基于基音周期模型的语音识别毕业设计（42页珍藏版）》请在毕设资料网上搜索。

1、 I 基于基音周期模型的语音识别 摘要 基音周期是表征语音信号本质特征的参数，属于语音分析的范畴，只有准确 分析并且提取出语音信号的特征参数，才能够利用这些参数进行语音识别处理。 语音识别率的高低，都依赖于对语音信号分析的准确性和精确性，因此基音周期 的研究在语音信号的处理应用中具有十分重要的作用。本论文通过两种算法在 Matlab 上实现基音周期的检测，分别是短时自相关函数法和短时平均幅度差函数 法。通过实验得到的基音周期结果的分析来识别不同的人的语音信号。 论文首先介绍了语音基音检测算法在语音识别方面的研究背景极其重要意 义。其次对现有的基音检测算法进行了归纳和总结，并详细的介绍本文将用的

2、两 种基本基音检测算法的基本原理及实现。最后在 Matlab 上对语音信号进行基音周 期的检测。 本设计为语音信号的基音周期检测，采集语音信号，对语音信号进行处理， 区分清音浊音，并通过对采样值进行滤波、分帧、求短时自相关函数，得到浊音 的基音周期。 关键词：预处理,基音检测,自相关函数法,平均幅度差函数法,基音轨迹 ，语 音识别 II Abstract pitch is characterization parameters of speech signal characteristics, belongs to the category of speech analysis, only t

3、he accurate analysis and extract the characteristic parameters of speech signal can use these parameters to speech recognition processing. voice recognition rate of high and low, are dependent on the accuracy of the analysis of speech signal and accuracy, so the pitch period of the study on speech s

4、ignal processing plays an important role in application.in this paper, through the two algorithm in Matlab pitch period detection, respectively is short-time autocorrelation method and short time average magnitude difference function method. Pitch period is obtained by the experiment results of spee

5、ch signal analysis to identify the different people. The curriculum design of pitch detection of speech signals, speech signal, the speech signal processing, distinguish between the voiced and unvoiced speech has different samples values in the filtering, framing, asks the short-time autocorrelation

6、 function, have voiced sound pitch period. The thesis first introduces the research background of speech pitch detection algorithm is very important. Next to the existing algorithm for pitch detection are summarized, and a detailed introduction to this article will use the two kinds of basic pitch detection algorithm is the basic principle and realization. Finally in Matlab on speech signal pitch period detection. Keywords: pretreatment,pitch detection ,autocorrel