A Study on the Filtering Method of Natural Gamma Energy Spectrum Logging Data for Low-Count Features

The natural radioactivity intensity of rocks is low, and the energy spectrum count measured by natural gamma spectrum logging is low. The low count characteristic magnifies the influence of random error of formation decay, resulting in poor quality of energy spectrum measurement and large statistical noise of logging curve. In this paper, a variety of spectral filtering results are compared, so that the single depth point spectral information can be restored to the maximum extent. By using the curve shape, depth domain filtering is carried out for the U, Th and K content curves to comprehensively extract formation information from the data with high noise and low count. The validity of the proposed method is verified by the measured data of the natural gamma-ray spectrometer of the drilling tool of China National Logging Corporation. The results show that the effect of Gaussian filter and convolution filter is relatively good for the natural gamma spectrum with low count and poor spectral shape, and the composite filter combining the two methods in this paper can accurately restore the accurate spectral shape. In depth domain filtering, Kalman filtering has some advantages over adaptive filtering. After energy spectrum filtering and depth filtering, the accuracy and repeatability of the results are greatly improved. The filtering method of natural gamma energy spectrum logging for low-count features introduced in this paper can effectively improve the measurement accuracy of U, Th and K.