A smoothing method for X-ray spectrum based on best beamforming

Abstract

X-ray spectrum is widely used in the field of analysis and measurement. Due to the effect of electronic noise and statistical fluctuation of particle, the measurement spectrum is unsmoothed. In order to improve the signal-to-noise ratio of the spectrum, many smoothing methods were applied to process spectrum. The conventional smoothing algorithms regard the spectra as time series. In fact, X-ray spectrum is an angle series, and its essential is spatial spectrum. So the conventional smoothing algorithms are not fit for smoothing X-ray spectrum. In this paper, a new measuring model for X-ray spectrum is built up, and a new smoothing method based on best beam forming is proposed for X-ray spectrum. It regards the measured X-ray spectrum as the classic spectrum estimation result of array signal. The spatial spectrum is re-estimated to smooth the spectrum by best beam forming. The relationship between cross-correlation coefficient and number of array cell is applied to choose parameter q. The experimental results show that the method is effective to smooth X-ray spectrum with high resolution when q is larger than 200.