Abstract：The paper introduced three common-used de-noise approaches by wavelet:wavelet decomposi-tion and reconstruction,threshold of wavelet trans-form and modular extreme of wavelet transform.The denoise processing is carried out by abovemen-boned approaches for simulated signals addingGauss white noise respectively, and correlation andanalysis of advantages and disadvantages of thesethree approaches showed that it is suitable to selectwavelet decomposition and reconstruction as de-noise approach that is characterized by simplemethod and rapid computational speed when thedenoise processing is carried out for the signalswith band-separated signal and noise and determi-nable noise;the threshold and modular extremeapproaches can be selected for the denoise process-ing of Gauss white-noise-bearing signals. Since thethreshold approach is characterized by obtainingthe approximate optimum evaluation of originalsignal。rapid computational speed and wide adapta-bility, which is most popular-used denoise approachamong the wavelet denoise methods. The modularextreme approach of wavelet transform has betterdenoise effects for the signals containing more sin-gularities, but its disadvantage is slow computa-tional speed. It should strike a balance between thedenoise effect and computational speed and effec-tively and reasonably select the denoise approachthat is suitable for the practical data when practi-tally using wavelet transform for denoise.
