Tracking of non-brownian particles using the Viterbi algorithm

摘要: We present a global tracking algorithm for particles with dynamic motion models. The augments existing track linking based on the Viterbi Gaussian Mixture Probability Hypothesis Density filter. This allows to use target velocities link tracks. can handle clutter, missed detections, and random appearance disappearance of in field view. also targets that switch between different models according Markov process. is evaluated synthetic datasets used ISBI 2012 Particle Tracking Challenge, which simulate vesicles, receptors, microtubules, viruses at particle densities signal noise ratios. evaluation shows our performs well across wide range problems both 2D 3D.