A novel approach to the localization and estimation of radioactivity in contaminated waste packages via imaging techniques

Dismantling nuclear power plants entails the production of a large amount of contaminated (or potentially contaminated) waste that must be disposed according to national and international regulations. A large part of the end products needs to be stored in special repositories, but a significant part of it is slightly contaminated or not contaminated at all, making it possible to free release it. One possible approach to free release measurements uses large clearance monitors (LCMs), chambers surrounded by plastic scintillation detectors that can measure up to 1000 kg of waste. Due to the composite nature of the detection system in a LCM, it is easy to imagine that one can apply 3D imaging algorithms to localize radioactive sources inside a waste package. In this work we will show how a special algorithm that maximizes the conditional informational entropy allows decisions about the clearance of portions of the sample.