Experimental verification of algorithms for detection and estimation of radioactive sources

The paper considers the problem of estimating the number of radioactive point sources that potentially exist in a designated area and estimating the parameters of these sources (their locations and strengths) using measurements collected by a low-cost Geiger-Müller counter. In a recent publication the authors proposed candidate algorithms for this task: the maximum likelihood estimator (MLE) and the importance sampling estimator based on progressive correction (PC) for source parameter estimation, and the minimum description length (MDL) for the estimation of the number of sources. Using real experimental data acquired during a recent radiological field trial in Pucka-punyal Military Area (Victoria, Australia), in the presence of up to three point sources of gamma radiation, this paper presents an experimental verification of the measurement model and algorithms proposed by us earlier. These experimental results show that while the MLE performs well when no more than two sources are present, the PC performs remarkably well for all data sets, which confirms our previous conclusions based on simulation studies alone.