Background Deduction Approach Based on Iterative Adaptive Window-Based Wavelet Transform and Gaussian Convolution Filtering for XRF

摘要

In the X-Ray fluorescence (XRF) analysis, valid characteristic peaks of elements are often superimposed on low-frequency backgrounds, which harm the qualitative and quantitative analysis of elements. Therefore, background deduction is an essential part of spectral pre-processing. To address the above issue, this paper proposes a new method for background deduction which combines iterative adaptive window-based wavelet transform (IAWWT) with iterative Gaussian convolution filtering (IGCF). To verify the validity of the proposed algorithm in this paper, XRF spectra of 59 soil standards are obtained. Firstly, background deduction is performed on the spectra by using IDWT, IAWWT, IGCF, and the method proposed in this paper. Secondly, the peak areas of Cr, Mn, and Cu in the spectra are linearly fitted to the true contents of the samples. The results show that the proposed method can effectively deduct the background, and the goodness of fit ( R^{2 ) of Cr, Mn, and Cu is better than the other three methods after background deduction. Among them, R2 of Cr increases to 0.99, R2 of Mn increases to 0.96, and R2 of Cu increases to 0.97.}