A closer look to probabilistic state estimation - case: particle filtering

Abstract

Particle Filter is a significant member of the group of methods aiming to provide reasonable solutions to the real-world problems by approximating the value of the posterior density function using probabilistic sampling. Particle filtering has been increasingly used by researchers for the last two decades with the advancements occurred in computational resources in order to solve such problems. This paper focuses on Particle Filtering in a way to be a complete tutorial for the beginner researchers by means of providing a quick theoretical framework of Particle Filtering in a step-by-step progressive manner starting with Bayesian Inference as well as providing a stimulating multi-target tracking example problem with solution.