Research Of Speech Enhancement Based On Wavelet Transforms

The main part of this thesis is speech enhancement based on wavelet transform. And when researching how to make a judgment between voiceless sounds and voiced sounds for signals, I found that there was something that could be used in the endpoint detection of noisy speech signals. So there is a charter for endpoint detection of noisy speech signals.The first part concentrates on speech enhancement. Speech enhancement is one of the most important parts of the speech signal processing. It plays an important role in speech recognition and speech coding. First this thesis introduce the different characteristics between speech, hearing and noise, discuss various speech enhancement methods, then do some research on the traditional spectral subtraction method. According to the theory of wavelet transform and the idea of spectral subtraction, a method on wavelet domain is presented in this thesis. This method can improve the SNR greatly, and remove the music residual noise effectively. The result shows this method performs better. But both spectral subtraction method and wavelet transform method damage voiceless sounds for signals. So a modified wavelet transform method is presented. After analyzing the character of speech, a judgment is made between voiceless sounds and voiced sounds for signals before speech enhancement. The effectiveness of the method is certified by computer emulation results.The second part discusses about the endpoint detection. The endpoint detection is basic technology of the speech signal processing. It also plays an important role in speech recognition, speech enhancement and speech coding. The previous methods can not perform well in low SNR circumstance such as the method based on short energy. In this thesis, we bring forth a new method based on wavelet transform. The experimental results show that this method performs better than that based on short energy in low SNR.