Advanced automatic analysis of Cloud Chamber images

Cloud chambers were among the rst particle detectors ever conceived; a few Nobel prizes have been awarded in the rst half of the past century to discoveries made by means of such device, most importantly the positron and the muon. A cloud chamber inspired by those original apparatuses but improved in several respects have been recently realized in our labs. The work proposed in the context of this thesis concerns with the analysis of the images gathered by the cloud chamber exposed to natural environmental radiation. In addition to the necessary tuning and optimization of the chamber and camera parameters, the primary goal is to set up an automatic particle identication technique, based on advanced machine learning classication algorithms. In the future unsupervised learning methods will be tested in order to categorize a few of the main particle track types automatically, i.e. without teaching the algorithm with labeled image examples.