Application Of Wavelet Analysis On Speech Processing

Owing to the rapid development of the computer and networking technology, multimedia play a significant role in it. The technology on Speech Processing, as a type of multimedia technology, has a tremendous potential and edge.Before the advent of Wavelet Analysis, the chief tool on Signal Processing is Fourier analysis. Fourier Analysis is an irreplaceable tool on processing stationary stochastic signals, but is at a loss on processing non-stationary time-varying signals that abound in real world. Thus Wavelet Analysis was introduced. It is development of methods on signal processing. During many years' theoretical and applied study, it is superior to other tools on processing time-varying signals.As a type of time-varying signals, it is inevitably a focus to apply Wavelet Analysis on speech processing. In this paper, through the meticulous study and practice of Wavelet Analysis and Speech Processing, the author explores some methods of applying Wavelet Analysis on Speech Processing. The author's work list as follows:* Applying Wavelet Analysis on the determination of pitch periodicity.* Applying Wavelet Analysis on the speech waveform coding.* Applying Wavelet Analysis on the speech hybrid coding.The study result shows that it is efficient to apply Wavelet Analysis on determining pitch periodicity and the experiment of applying Wavelet Analysis on speech coding basically achieves our goals.All programs in this paper are implemented in Matlab5.2.