【正文】
攀枝花學院本科畢業(yè)設計 (論文 ) 摘要 I 摘要 基音周期是表征語音信號本質(zhì)特征的參數(shù),屬于語音分析的范疇,只有準確分析并且提取出語音信號的特征參數(shù),才能夠利用這些參數(shù)進行語音識別處理。語音識別率的高低,都依賴于對語音信號分析的準確性和精確性,因此基音周期的研究在語音信號的處理應用中具有十分重要的作用。本論文通過兩種算法在Matlab 上實現(xiàn)基音周期的檢測,分別是短時自相關函數(shù)法和短時平均幅度差函數(shù)法。通過實驗得到的基音周期結(jié)果的分析來識別不同的人的語音信號。 論文首先介紹了語音基音檢測算法在語音識別方面的研究背景極 其重要意義。其次對現(xiàn)有的基音檢測算法進行了歸納和總結(jié),并詳細的介紹本文將用的兩種基本基音檢測算法的基本原理及實現(xiàn)。最后在 Matlab 上對語音信號進行基音周期的檢測。 本設計為語音信號的基音周期檢測,采集語音信號,對語音信號進行處理,區(qū)分清音濁音,并通過對采樣值進行濾波、分幀、求短時自相關函數(shù),得到濁音的基音周期。 關鍵詞 : 預處理 ,基音檢測 ,自相關函數(shù)法 ,平均幅度差函數(shù)法 ,基音軌跡 ,語音識別攀枝花學院本科畢業(yè)設計(論文) Abstract II Abstract pitch is characterization parameters of speech signal characteristics, belongs to the category of speech analysis, only the 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 signal processing plays an important role in this paper, through the two algorithm in Matlab pitch period detection, respectively is shorttime autocorrelation method and short time average magnitude difference function method. Pitch period is obtained by the experiment results of speech 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 shorttime autocorrelation 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 ,autocorrelation function。 the average magnitude difference function method。 pitch contrail, Speech recognition 攀枝花學院本科畢業(yè)設計(論文) 目錄 目 錄 摘 要 .................................................................................................................................... I Abstract ............................................................................................................................ II 1 緒論 ............................................................................................................................. 1 ........................................................................................................................1 ..............................................................................2 ...................................................................................................3 .......................................................................................................4 2 基音檢測常用的算法 .................................................................................................. 5 ........................................................................................................................5 常用的基音檢測算法及其原理 .................................................................................5 自相關函數(shù)法 ................................................................................................5 平均幅度差函數(shù)法 .........................................................................................8 本章小結(jié) .............................................................................................................. 10 3 基于 Matlab 基音周期的檢測 ...................................................................................11 語音文件的錄制與讀入 ......................................................................................... 11 語音信號的采樣和分幀 ......................................................................................... 12 語音信號基音周期的檢測 ...................................................................................... 16 短時自相關函數(shù)法檢測基音周期 .................................................................. 16 短時平均幅度差函數(shù)法檢測基音周期 ........................................................... 17 基音周期檢測的后處理 ......................................................................................... 19 濾波后的基音周期檢測 ......................................................................................... 23 本章小結(jié) .............................................................................................................. 25 4 語音識別結(jié)果分析 .................................................................................................... 26 5 展望與應用 ................................................................................................................. 30 附錄 ................................................................................................................................. 33 致 謝 ............................................................................................................................. 39 攀枝花學院本科畢業(yè)設計(論文) 1 緒論 1 1 緒論 引言 通過語言相互傳遞信息是人類最重要的基本功能之一。語言是人類特有的功能,它是創(chuàng)造和記載幾千年人類文 明史的根本手段,沒有語言就沒有今天的人類文明。同時語言又屬于聲學范疇,是相互傳遞信息最重要的手段,是人類最重要,最有效,最常用和最方便的交換信息的形式。語音中除了包含實際發(fā)音內(nèi)容的語音信息外,還包括發(fā)音者是誰及喜怒哀樂的各種信息。語音識別技術是一項集聲學,語音學,計算機,信息處理,人工智能計算機接口的范疇?,F(xiàn)在的電腦網(wǎng)絡技術的普及和完善,對于語音信號處理技術發(fā)揮著越來越重要的作用。語音識技術是用來設計一種會聽話的計算機,只要對著麥克風講話,便可以指揮計算機動作,也就是要實現(xiàn)“芝麻開門”聲控計算機的夢想,當然 現(xiàn)在這已經(jīng)不是夢想。語音信號處理的一系列技術及其應用已經(jīng)成為信息社會中不可或缺的組成部分,促進了信息社會的發(fā)展。語音信號中的情感信息是一種很重要的信息資源,它是人們在感知事物時的一