Bayesian path estimation using the spatial attributes of a road network
Mark R. Morelande, Matt Duckham, Allison Kealy, Jonathan Legg · 2015
We consider the problem of estimating the path taken by an object in a road network from sparse, noisy position measurements. Path estimation is posed in a Bayesian framework which allows the incorporation of prior information about vehicle movements. A carefully designed importance sampler is used to approximate the posterior path probabilities. The algorithm is demonstrated on simulated data.