Wavelet Thresholding and Sliding Mean Filtering-Based Signal Denoising for Atomic Fluorescence Spectroscopy

Atomic fluorescence spectral signals usually contain noise, and the noise will affect the analysis of the signals, so it is necessary to process the spectral signals. Aiming at the characteristics of atomic fluorescence spectral signals, this paper proposes to denoise the signals by using wavelet transform combined with sliding mean filtering, which first denoises the signals by wavelet thresholding and then filters the signals by sliding mean filtering. The signal-to-noise ratio (SNR), mean square error (MSE), and waveform similarity parameter (NCC), are used to evaluate the effectiveness of spectral signal processing, and the final experimental results show that the hybrid filtering method of wavelet threshold combined with moving mean can effectively remove the noise in the atomic fluorescence spectrum signal.