A comparison of methods for baseline estimation in gamma-ray spectrometry

Abstract Four programs for the estimation of the baseline of gamma-ray spectra were tested for accuracy of estimation in baseline regions, preservation of peak location and net area in baseline subtraction, precision and program size and speed. Programs GAMBAK and FILBAK performed better than programs AVBAK and BACKGR in estimating baselines between peaks. The four programs performed similarly with respect to accuracy, with GAMBAK producing the best results. FILBAK and GAMBAK estimated baselines under peaks with better precision than did AVBAK and BACKGR. GAMBAK, BACKGR and FILBAK were similar in size and speed while AVBAK was smaller but much slower.