.png){kind=icon}
.png){kind=icon}
Return
Measurement based dimension descent association algorithm for OTHR multi-detection multi-target tracking
.png){kind=icon}
PRE
AI
DOI：10.1016/j.isatra.2023.05.016.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The problem of multi-detection multi-target tracking (MDMTT) using over-the-horizon radar in dense clutter environment is studied in this paper. The biggest challenge of MDMTT is the 3-dimensional multipath data association among measurements, detection models and targets. In particular, a lot of clutter measurements are generated in dense clutter environment, which increase the computational burden of 3-dimensional multipath data association greatly. A measurement based dimension descent association (DDA) algorithm is proposed to solve the 3-dimensional multipath data association, which decomposes the 3-dimensional multipath data association into two 2-dimensional data asso-ciations. The proposed algorithm can reduce the computational burden compared with the optimal 3-dimensional multipath data association and the computational complexity is analyzed. Besides, a time extension method is designed to detect the new-born targets that appear in the tracking scene, which is based on the sequential measurements. The convergence of the proposed measurement based DDA algorithm is analyzed. The estimation error can convergence to 0 as the number of Gaussian mixtures tends to infinity. The effectiveness and rapidity of the measurement based DDA algorithm are demonstrated by the comparative simulation with the previously proposed algorithms.(c) 2023 ISA. Published by Elsevier Ltd. All rights reserved.
Keywords:
Dimension descent association
Target detection
Convergence analysis
Multi-detection multi-target tracking