Fusion of Range and Luminance Data from Laser Radar Systems.

摘要

In this project, a technique was developed for computing the optimal description of some physical scene by using global optimization and Bayesian techniques to combine correlated outputs of diverse nonlinear sensors. Two laser range/intensity imagers were considered, but the major application was to obtain the best known restorations of very noisy PD-, TI-, and T2-weighted Magnetic Resonance Images by fusing the information in all the input images. Sensor fusion, Magnetic resonance imaging, Mean-field annealing.