Reliable Hybrid Mixture Model for Generalized Point Set Registration

Point set registration (PSR) is an essential problem in the field of surgical navigation and augmented reality (AR). In navigation, aim mapping pre-operative space to intra-operative space. This article introduces a reliable hybrid mixture model, which reliability normal vectors generalized point (GPS) examined exploited. The motivation considering orientation information that cannot be estimated or measured accurately clinic. (PS) divided into two subsets according vectors. PSR cast maximum likelihood estimation (MLE) problem. expectation maximization (EM) framework used solve MLE E-step, posterior probabilities between points PSs are computed. M-step, transformation matrix model components updated by optimizing objective function. We have demonstrated through extensive experiments on human femur bone PS proposed algorithm outperforms state-of-the-art ones terms accuracy, robustness, convergence speed.