Study On The Background Deduction Method In Gamma-ray Spectrometry Data And Software Development

The background of NaI(Tl) gamma-ray spectral data is very complex. The affecting factors include cosmic radiation, environmental background and Compton scattering effect produced by the interaction between gamma-ray and other material. All gamma-ray characteristic peaks are added on the background. Therefore the separation of all peaks from the background is a key technology. In this paper, based on the conventional straight-line method, the background reduction methods using Fourier transform and the peak stripping are studied. The processed results are compared and the realization software is developed.The Fourier transforms method implements background reduction in gamma-ray spectral data by means of Fourier transform. The gamma-rays background spectoum is a continuous smooth curve and can be perceived as low-frequency components. In the data processing, characteristic peaks can be perceived as the useful signals added on the smooth line of the background, and as components of high-frequency. The initial spectrum is produced through the Fourier transform. Low-frequency components are filtered by an ideal filter while the characteristic peaks of high frequency information are retained. Thus the background is reduced. Peak stripping method is based on the spin-off spectrum rapid changes in the characteristic signals. The information of one channel is compared with that of its neighboring channel and the smaller one is taken. The transformation is run in all channels. Thus the amplitude decreased at the locations of the peaks while the amplitude at the other locations keeps unchanged. As a result of this method, the two local minimum tend to be connected, resulting in smooth spectrum. The shape of the background is gained after several times of stripping.In software design, based on VC + +6.0 visual programming software interface, image display, data pre-processing, the background reduction and the results output program are completed,. Functions and compiling process of the software are described in detail and the key codes are provided in the thesis.The results demonstrate that the methods using Fourier transform and peak stripping can effectively separate and reduce the background. At the same time, the methods also extract information of a number of characteristic peaks. In conclusion, the methods are superior to straight-line method on background reduction. The designed software makes data processing simple, visual, easy-to-use and helpful in comparing and analyzing each element of the processed results. The methods have practical value.