Due to the multiscale character of wavelet transform, the method of wavelet transform de-noising (WTDN) is introduced. The WTDN method neither requires extra limit of frequency range for the processed signal, nor needs a prior estimate of impulse response for an identified system, so it is especially suitable to de-noise the wide-band signal and the impulse response of a blind system. The numerical simulation results indicate that the WTDN method is reliable. The WTDN method was used to process the sampled data from a preamplifier coupled to a gas detector. The experimental results also show that the WTDN method can effectively improve the SNR of sampled data and enhance the accuracy in pole identification of the system.