RICH pattern recognition

An algorithm for performing pattern recognition in RICH detectors is described. It uses the knowledge of particle trajectories through the detector, along with the optics of the RICH, to predict where photons should be expected on the photodetector plane, for a given choice of particle mass-hypotheses. This prediction is compared to the observed distribution of detected photoelectrons, and a likelihood is calculated. The maximum-likelihood solution is then search for. The algorithm has been applied to a simulation of the RICH system of LHCb, and its performance is described.1