PMBM-Based Unresolved-Group Object Tracking
Guchong Li, Gang Li, You He
Abstract
The focus of this paper is to tackle the challenging task of unresolved-group object (UO) tracking by exploiting the Poisson multi-Bernoulli mixture (PMBM) filter, named UO-PMBM. Specifically, according to the UO likelihood function, the probability generating functional (PGFL) tool and functional derivative are first used to derive the filtering recursion expressions of the UOPMBM. Then, detailed descriptions of the Gaussian mixture (GM) implementations are described. Lastly, the effectiveness of the proposed UO-PMBM approach is demonstrated through simulation experiments.
Topics & Concepts
Radar trackerComputer scienceGroup (periodic table)Object (grammar)Computer visionArtificial intelligenceTracking (education)RadarPhysicsTelecommunicationsQuantum mechanicsPsychologyPedagogyVideo Surveillance and Tracking Methods