Particle-inspired motion updates for grid-based Bayesian trackers

The computational cost of the motion update has limited the application of grid-based Bayesian trackers. Drawing inspiration from particle filters, an algorithm for efficient grid-based motion updates is developed. The algorithm's complexity is linear in the number of grid cells and independent of the time increment for the motion update. It has the flexibility to model any Markov motion process. The accuracy of the algorithm and its sensitivity to implementation parameters is assessed, and trade-offs between accuracy and computational cost are explored.