Sensitivity Improvement In Low-Profile Distributed Detector Systems For Tracking Sources In Transit

The RadTrac real-time detection and tracking software runs on a laptop computer networked to gamma-radiation detectors. A probabilistic estimate for source position is generated by combining measured count rate data with a first-principles stochastic model for the space and time dependence of count rates and knowledge of detector intrinsic efficiency. Recent development work has focused on improving RadTrac sensitivity in low-count rate situations. A method has been developed for processing count rates by energy according to that part of the energy spectrum with the greatest signal-to-noise ratio. In addition a method has been developed that places constraints on the solution that are physically appropriate when count rates approach background. In both instances experiments with a weak source confirmed the uncertainty in estimated position is reduced.