【正文】 Matlab 語音信號處理 摘要 語音信號處理是研究用數(shù)字信號處理技術(shù)和語音學(xué)知識對語音信號進(jìn)行處理的新興的學(xué)科，是目前發(fā)展最為迅速的信息科學(xué)研究領(lǐng)域的核心技術(shù)之一。通過語音傳遞信息是人類最重要、最有效、最常用和最方便的交換信息形式。 Matlab語言是一種數(shù)據(jù)分析和處理功能十分強(qiáng)大的計算機(jī)應(yīng)用軟件，它可以將聲音文件變換為離散的數(shù)據(jù)文件，然后利用其強(qiáng)大的矩陣運(yùn)算能力處理數(shù)據(jù)，如數(shù)字濾波、傅里葉變換、時域和頻域分析、聲音回放以及各種圖的呈現(xiàn)等，它的信號處理與分析工具箱為語音信號分析提供了十分豐富的功能函數(shù)，利用 這些功能函數(shù)可以快捷而又方便地完成語音信號的處理和分析以及信號的可視化，使人機(jī)交互更加便捷。信號處理是 Matlab重要應(yīng)用的領(lǐng)域之一。 本設(shè)計通過下載 一段語音，對其進(jìn)行了時域分析，頻譜分析 ,分析語音信號的特性。并應(yīng)用 matlab平臺對語音信號 除去噪聲，進(jìn)一步設(shè)計 三 種濾波器即高通濾波器、低通濾波器 、 帶通濾波器 ， 基于洪澤 三 種濾波器設(shè)計原理，用窗函數(shù)法設(shè)計出這 三 種濾波器，然后含加噪的語音信號 進(jìn)行濾波處理。最后對比濾波前后的語音信號的時域和頻域特性，回放含 噪語音信號和去噪語音信號，對比研究了IIR 濾波器和 FIR濾波 器在信號處理尤其是信號濾波去噪方面的特性和應(yīng)用。論文從理論和實(shí)踐上比較不同數(shù)字濾波器的濾波效果。 關(guān)鍵字： Matlab；語音信號 ； 信號處理；傅里葉變換；濾波器 The Design of Analysis Voice Signal Abstract Speech signal processing is to study the use of digital signal processing technology and knowledge of the voice signal voice processing of the emerging discipline is the fastest growing areas of information science one of the core technology. Transmission of information through the voice of humanity39。s most important, most effective, most popular and most convenient form of exchange of information.. Matlab language is a data analysis and processing functions are very powerful puter application software, sound files which can be transformed into discrete data files, then use its powerful ability to process the data matrix operations, such as digital filtering, Fourier transform, when domain and frequency domain analysis, sound playback and a variety of map rendering, and so on. Its signal processing and analysis toolkit for voice signal analysis provides a very rich feature function, use of these functions can be quick and convenient features plete voice signal processing and analysis and visualization of signals, makes puter interaction more convenient . Matlab Signal Processing is one of the important areas of application. The design of mobile phone collection a noisy speech, has carried on the analysis in time domain, frequency spectrum analysis, analysis of the characteristics of speech signal. And application of the MATLAB platform to the voice signal to remove noise, the further design of three kinds of filter is a highpass filter, lowpass filter, bandpass filter, based on the Hongze three kinds of filter design, window function method to design the three kinds of filter, and then with the original speech signal filtering. Finally the parison before and after filtering the speech signal in time domain and frequency domain, the playback noise speech signal denoising and the voice signal, a parative study of the IIR filter and a FIR filter in signal processing especially for signal denoising in terms of the characteristics and application of. The theory and practice from the parison of different digital filtering effect. Keywords: Matlab。 speech signal。 signal processing。 FFT。 filter 目錄 摘要 ................................................................................. I Abstract ............................................................................ II 第一章緒論 ........................................................................... 1 ................................................................ 1 ................................................................ 1 ...................................................... 2 .......................................................... 2 .......................................................... 2 第二章語音信號處理的總體方案 .......................................................... 4 ...................................................... 4 .............................................................. 4 .......................................................... 4 .................................................... 4 Wavread函 數(shù)介紹 ........................................................ 4 ........................................................ 5 第三章語音信號處理的基本知識 .......................................................... 6 .............................................................. 6 ............................................................ 6 時域信號 FFT 分析 ............................................................ 6 第四章 語音信號的采集 ................................................................ 7 ................................................................ 7 ............................................................ 8 第五章 數(shù)字濾波器的設(shè)計和畫出其頻率響應(yīng) .............................................. 10 ......................................................... 10 FIR濾波器的性能比較 .............................................. 10 FIR濾波器 ............................................................... 11 ...................................................... 11 ...................................................... 12 的設(shè)計 .................................................... 13 IIR濾波器 ............................................................... 14 .................................................... 14 ...................................................... 15 ...................................................... 17 第六章 語音信號的濾波 ............................................................... 19 ........................................................... 19 ......................................................... 19 ..................................................... 19 ..................................................... 21 第 七 章總結(jié) .......................................................................... 25 參考文獻(xiàn) ............................................................................ 26 致謝 ................................................................................. 1 附錄： ............................................................................... 1 第一章緒論 課題研究的背景 通過語音傳遞倍息是人類最重要 、最有效、最常用和最方便的交換信息的形式。語言是人類持有的功能， 聲音是人類常用的工具，是相互傳遞信息的最主要的手段。因此，語音信號是人們構(gòu)成思想疏通和感情交流的最主要的途徑。并且，由于語言和語音與人的智力活動密切相關(guān)，與社會文化和進(jìn)步緊密相連，所以它具有最大的信息容量和最高的智能水平。現(xiàn)在，