Optimal Scheme for Detecting Weak Peak in γ-ray Spectrum

A peak search method based on spectrum convolution was proposed. Required preconditions, by which exact position of weak peaks would be detected, were analyzed. A suit of optimal scheme for detecting all of weak peaks in γ-ray spectrum was worked out. It contained a filter with noise level detected by standard deviation, background eradication by SNIP algorithms, and removing false peaks and resolving overlapping peaks by detection limit and the size of opposite peak width. All of programs about this scheme and two other peaks search algorithms were compiled in Matlab 7.1 platform. Test spectrum of IAEA in 1976 was analyzed by the above three and two commercial γ spectrum analysis software. The results show that this scheme is preferable to the others. This method can distinguish available Compton platform and peaks, and check out all of the 22 peaks without a false peak. The largest deviation between calculated position and standard position by this method is 2.90 channels.