Abstract:In practical target tracking applications, it is challenging to accurately establish the measurement noise covariance model. The traditional Probability Hypothesis Density filtering algorithm's tracking performance rapidly deteriorates when measurement noise is unknown. To address the impact of unknown measurement noise on multi-extended target tracking, Variational Bayesian methods are introduced into the Probability Hypothesis Density filtering algorithm. The Variational Bayesian Probability Hypothesis Density algorithm allows filtering calculations by estimating noise covariance in cases of unknown measurement noise. However, this algorithm still suffers from inaccurate target number estimation. To solve this problem, this paper proposes an improved Variational Bayesian Probability Hypothesis Density algorithm. This algorithm introduces elliptical thresholds after the prediction step in the filtering process and utilizes the measurements retained within the thresholds for further calculation, in order to reduce the influence of clutter measurements on the extended target measurement, and improve the precision of the extended target state clustering. Experimental results show that the proposed algorithm is suitable for the situation where the number of multiple extended targets is unknown and the measurement noise covariance is unknown, and its tracking accuracy is better than the traditional probability hypothesis density and Variational Bayesian probability hypothesis density filtering algorithms.

Multiple Gaussian Processes Based Extended Target Tracking with Variational Inference

Irregular Extended Target Tracking with Unknown Measurement Noise Covariance

Single Target Tracking with Unknown Detection Status: Filter Design and Stability Analysis

Novel multiple-model probability hypothesis density filter for multiple maneuvering targets tracking

Variation Bayesian Interference for Multiple Extended Targets or Unresolved Group Targets Tracking

Multiple Extended Target Joint Tracking and Classification Based on GPs and LMB Filter

A Gaussian Process Regression based Dynamical Models Learning Algorithm for Target Tracking

Collaborative penalized Gaussian mixture PHD tracker for close target tracking.

Visual Tracking Via Locally Structured Gaussian Process Regression

The Multiple Model Poisson Multi-Bernoulli Mixture Filter for Extended Target Tracking

Multi-extended Targets PHD Filtering Tracking Algorithm with Unknown Measurement Noise Distribution

Bounding Multiple Gaussians Uncertainty with Application to Object Tracking

Episodic Gaussian Process-Based Learning Control with Vanishing Tracking Errors

Non-Ellipsoidal Infrared Group/Extended Target Tracking Based on Poisson Multi-Bernoulli Mixture Filter and B-Spline

Monocular Tracking 3D People by Gaussian Process Spatio-Temporal Variable Model

A Gaussian Mixture CPHD Filter for Multi-Target Tracking in Target-Dependent False Alarms

A Multiple Hypothesis Tracking Method for Extended Target Tracking

Linear Time GPs for Inferring Latent Trajectories from Neural Spike Trains

Large Linear Multi-output Gaussian Process Learning

The Gaussian Particle multi-target multi-Bernoulli filter

Multiple-Target Tracking by Spatiotemporal Monte Carlo Markov Chain Data Association