Graph-Based Multiobject Tracking with Embedded Particle Flow

TL;DR

This paper introduces a graph-based Bayesian multiobject tracking method that embeds particle flow to improve detection and tracking efficiency in high-dimensional scenarios with fewer particles.

Contribution

It presents a novel graph-based Bayesian approach with embedded particle flow for nonlinear, high-dimensional multiobject tracking, reducing computational load.

Findings

Reduced computational complexity and memory usage

Effective tracking with fewer particles in high-dimensional scenarios

Favorable detection and estimation accuracy in 3-D scenarios

Abstract

Seamless situational awareness provided by modern radar systems relies on effective methods for multiobject tracking (MOT). This paper presents a graph-based Bayesian method for nonlinear and high-dimensional MOT problems that embeds particle flow. To perform operations on the graph effectively, particles are migrated towards regions of high likelihood based on the solution of a partial differential equation. This makes it possible to obtain good object detection and tracking performance with a relatively small number of particles even if object states are high dimensional and sensor measurements are very informative. Simulation results demonstrate reduced computational complexity and memory requirements as well as favorable detection and estimation accuracy in a challenging 3-D MOT scenario.