Robust extended Cubature Kalman multi-Bernoulli filter for joint detection and tracking ISAR images of small UAVs

ABSTRACT

A novel approach to joint detection and tracking the moving extended small vehicles such as drones and unmanned aerial vehicles (UAVs) using inverse synthetic aperture radar (ISAR) images is proposed in this work. Recently, a random matrices model (RMM) for extended object tracking (EOT) has been proposed. Under the RMM, the object shape is assumed to be an ellipse. The RMM is chosen to describe the EOT shape because it was a good combination of an informative shape model and low computational complexity. Since conventional EOT algorithm using RMM is not accurate enough to reconstruct the object’ extension by one ellipse, an appropriate technique needs to be developed for ISAR imaging and tracking of extended drones tracking (EDTs). Therefore, we applied a new tracking-before-detection (TBD) algorithm for EDT as multiple ellipses or sub-RMM with non-linear ISAR observations and unknown orientation angle due to Doppler effect. In the recent year, the Cubature Kalman multi-Bernoulli (CK-MB) algorithm with a third-degree spherical-radical cubature rule has been applied to solve the non-linearity due to Doppler effect, the CK-MB filter is more accurate compared to state-of-the-art algorithms such as sequential Monte Carlo (SMC)-MB and Gaussian mixture (GM)-MB filters. An approach to joint detection and tracking of moving EDTs using ISAR images with the robust extended CK (ECK)-Sub-RMM-MB-based TBD algorithm is proposed and evaluated in this paper. We describe the EDT state through the ECK-Beta Gaussian inverse Wishart model. Simulation results demonstrates this remarkable performance.

KEYWORDS:

• ISAR
• cubature Kalman MB-TBD
• extended vehicles tracking
• Doppler-only measurements
• sub-RMM

Disclosure statement

No potential conflict of interest was reported by the author(s).