基于MATAB的时域和频域分析语音进行数字处理.docVIP

摘 要 语音信号处理包括语音通信、语音增强、语音合成、语音识别和说话人识别等方面。只有通过语音信号的数字处理，语音信号的好坏、语音识别率的高低，都取决于语音信号处理的好坏。因此，语音信号处理是一项非常有意义的研究课程。 语音，即语言的声音，是语言符号系统的载体。它由人的发音器官发出，负载着一定的语言意义。语言依靠语音实现它的社会功能。语言是音义结合的符号系统，语言的声音和语言的意义是紧密联系着的，因此，语言虽是一种声音，但又与一般的声音有着本质的区别． 关键词：语音；噪声；MATLAB；滤波器；信号处理 Abstract Speech signal processing including voice communications, speech enhancement, speech synthesis, speech recognition and speaker recognition and so on. Only through the digital processing of voice signals, voice signals good or bad, high and low speech recognition rate are good or bad depends on the voice signal process .Therefore, the voice signal processing is a very significant research programs. Voice, the voice of the language is the carrier of linguistic sign systems. It is issued by the human vocal organs, load a certain amount of linguistic meaning. Language dependent voice to achieve its social function. Language is the combination of sound and meaning of the symbol system, language, the meaning of the voice and language are closely linked, and therefore, although the language is a voice, but the sound in general is essentially different. Voice that people exchange ideas and conduct the most basic means of social activities, so we have to deal with the speech signal analysis and optimization of human communication exchange. The training record a tape recorder with windows own words (audio signal), and draw the speech signal in time domain waveform and frequency spectrum; In the MATLAB software platform, to the speech signal with noise, drawn after the speech signal plus noise time domain waveform and frequency spectrum; Bilinear transform method, to design a filter, and draw the frequency response filter; Designed to filter their own filtering of the speech signal, draw the filtered signal in time domain waveform and frequency spectrum, and filter before and after the signals are compared to draw difference time-domain waveform analysis language signal changes before and after filtering . The trai