The Study Of Method Based Wavelet Transform In Removing Noise And Its Application

Signal and information processing is one of the most raPidly developedsubject in the field of information science in recent ten years. Typical signalprocessing mechanism, such as pure time domain, pure frequency domain, Shortterm Fourier transform, wigner-ville distribution and so on have their ownlimitations. As the result, their aPplication eXtent are restricted. WaVelettransform, which developed in latel980's, is a new kind of time scale and timefrequency analytic method. It has good localization characteristic in both timeand frequency domain. Nowadays, it is widely used in practical aPplications.This thesis will analyze, compare and discuss the separate aPplications on differentkinds of typical remove noise processing methods. In addition, focus on WavelettransfOrm theory, we analyze its characteristic and take its advantage tO implement anew remove noise mechanism. Both theory analyzing and experimefit result on non-stationary signal in order to certify the WaVelet transform remove noise method issuperior to the typical ones. Detection on signal singularity point is also discussed inthis thesis. WaVelet transform can locatC the signal singularity point precisely, however,Fourier transfOrm can only flnd the global singularity. Obviously, it is advantage overtypical methods.