Due to the limitations of the present artificial signal recognition technology and Fourier transform in analyzing the mine microseismic signal, the Matlab wavelet toolbox analysis method was presented. Through the transformation of the microseismic signals monitored in mine, prior to the spectral analysis of transformation signals, the discontinuity of the signals were accurately observed and the first arrival time of P wave was determined. As a result, the accurate location and energy of mining shocks were concluded. Further, through the wavelet de-noising of signals and comparing the four kinds of de-noising methods, the unbiased estimate threshold method worked best and this showed the powerful function of the wavelet analysis in de-noising. Therefore, it is revealed that the wavelet analysis is an effective method of seismic signal processing and analysis as it is capable of seismic signal recognition and noise elimination.