Study for wavelet denoising methods of laser Raman spectrum of carbendazim pesticide

Raman signal of carbendazim pesticide was collected by laser Raman spectrometer. Moving- average smoothing method, wavelet soft- threshold method and wavelet hard- threshold were applied respectively to denoise the acquired Raman signals, and their denoising effects were compared.The results showed that wavelet hard- threshold could obtain the optimal denoising quality. When wavelet base function being sym2, scale decomposition being 5, the threshold quantization being ‘Heursure', hard- threshold value being processed, signal to noise ratio ( SNR) of the reconstructed spectral was maximum which was 60.927, and root mean square error ( RMSE) was minimum which was 11.429. The study showed that wavelet hard- threshold could effectively remove the noise information of the raw Raman signals of carbendazim pesticide, and saved the most number of spectra details.The study could provide a methodological support for the rapid detection of pesticide residue in food and agricultural products based on Raman spectrum.