Application Of Wavelet Transform In Denoising The Nal γ Energy Spectrum

Nal detectors are widely used in channel monitoring, radiation protection, environmental protection and other fields. During the NaI (Tl) spectrum measurement, the instrument itself, the external environment, and human factors etc., that inevitably contain various noise, will influence the results,so that the resulting spectrum data, these higher levels of noise often limits the application of Nal detector. Therefore, the need to reduce noise and improve signal to noise ratio is very important. Noise reduction is the maximum retention of the original signal of the measured data while minimizing noise. For it has better localization properties in space and time domains than classic signal noise reduction ways,wavelet transform has a wide range of applications in signal noise reduction.This major work is to study the basic theory of wavelet transform and use their spectrum for Nal measured signal noise analysis specific tasks are as follows:First, the basic concepts of wavelet transform is discussed, such as:Fourier transform, short-time Fourier transform, continuous wavelet transform, discrete wavelet transform, multiresolution analysis theory and MALLAT algorithm.The137Cs measured spectrum is analysed by the db4wavelet hard threshold and soft threshold denoising method, respectively, using Matlab. The result shows that wavelet soft threshold denoising method is a better way to analyse Nal (Tl)γ spectrum.To determine the appropriate sampling channel numbers, the same object, different number of channels measured spectra is analyzed by wavelet noise reduction.Finally, according to previous work, a mixed spectrum of137Cs and60Co which conducted noise reduction peak area, peak Compton platform ratio is analyzed. It proves that wavelet transform noise reduction is effective to improve Nal spectrum analysis.