No abstract

Assuming that the measurement and process noise covariances are known, the probability hypothesis density (PHD) filter is effective in real-time multi-target tracking; however, noise covariance is often unknown and time-varying for an actual scene. To solve this problem, a strong tracking PHD filter based on Variational Bayes (VB) approximation is proposed in this paper. The measurement noise covariance is described in the linear system by the inverse Wishart (IW) distribution. Then, the fading factor in the strong tracking principle uses the optimal measurement noise covariance at the previous moment to control the state prediction covariance in real-time. The Gaussian IW (GIW) joint distribution adopts the VB approximation to jointly return the measurement noise covariance and the target state covariance. The simulation results show that, compared with the traditional Gaussian mixture PHD (GM-PHD) and the VB-adaptive PHD, the proposed algorithm has higher tracking accuracy and stronger robustness in a more reasonable calculation time.

No abstract

No abstract

A new variational Bayesian-based adaptive distributed fusion unscented Kalman filter (ADFUKF-VB) algorithm is proposed for the problem of state estimation in a distributed fusion target tracking system with unknown sensor measurement losses. First, the Bernoulli random variables are introduced as the judgement factors of the measurement losses in each local unscented Kalman filter (UKF) of the distributed fusion target tracking system, and the probability density function (PDF) of the measurement loss probability is modelled as a Beta distribution. Then, the variational Bayesian (VB) approach is used to infer the estimated variables in each local UKF. The measurement losses are judged according to the expectations of the Bernoulli random variables, and three ways are adopted to estimate the state vector based on the obtained judgement results. Finally, the covariance intersection (CI) fusion strategy is used to achieve the fusion of all local state estimates. The simulation results show that the proposed ADFUKF-VB algorithm can accurately judge whether the measurement is lost and estimate the unknown measurement loss probability, moreover, effectively improve the accuracy of state estimation compared with the existing filtering algorithms with unknown measurement losses. K E Y W O R D S sensor fusion, signal processing, state estimation, target trackingThis is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.