Millimeter-wave radar has been widely adopted in intelligent transportation systems. Modern millimeter-wave radars' high resolution makes a single target yield multiple measurements (point cloud measurements), turning target tracking into an extended target tracking (ETT) problem. The uncertainty in radar measurement sources coupled with the complex spatial distribution of measurements fundamentally challenges ETT algorithms. Conventional ETT algorithms with cluster-then-associate frameworks partition point cloud measurements predominantly through density, failing to exploit target shape information and resulting in suboptimal clustering efficacy. This leads to inaccurate associations between measurements and extended targets, ultimately degrading overall tracking accuracy. This paper proposes a novel approach to circumvent the performance limitations caused by clustering errors in traditional methods. First, we propose a clustering-free closed-loop ETT framework that incorporates the prior target's shape information as feedback. Subsequently, we develop a data association method that leverages the inherent correlation between point cloud measurements and target shape. The associated measurements are then probabilistically fused and integrated into a Kalman filter for state updating. In simulated and real-world datasets, compared with the traditional method, clustering before association, we have validated the effectiveness of the proposed method.

With the improvement of sensor resolution, extended targets (ETs), characterised by multiple measurements, have become a hotspot in the field of target tracking. The trajectory probability hypothesis density (TPHD) filter, which inherently provides trajectory estimates, stands out as an effective tracking filter. By integrating the interacting multiple model (IMM) approach, we introduce an IMM-ET-TPHD filter for tracking maneuvering ETs in this paper. Firstly, we extend the concept of the IMM to trajectory sets and derive the IMM-ETTPHD recursion. Then, we develop a Gaussian mixture (GM) implementation of the IMM-ET-TPHD recursion. Finally, the proposed IMM-ET-TPHD algorithm is validated in a challenging maneuvering ET tracking scenario, demonstrating its superior performance.

Joint tracking and classification (JTC) of vehicles is a crucial yet challenging task in intelligent transport and automotive systems. The advent of high-resolution modern sensors necessitates treating vehicles as extended targets. Current extended target tracking (ETT) algorithms provide shape estimations for vehicles, making shape size the most intuitive and accessible feature for classification. This paper contributes two key elements to achieve the JTC of vehicles. For one thing, we introduce the rectangular constraints and customize distinguishable measurement models using modified Gaussian processes (GP). For another thing, based on the customized GP models, we strengthen the role of class in the conditional extended target probability hypothesis density (ET-PHD) filter. Subsequently, we propose a class-enhanced JTC-ET-PHD filter and its Gaussian mixture implementation, enabling simultaneous kinematic, shape, and class estimation of vehicles. Finally, numerical results validate the proposed shape estimation and JTC method, affirming their effectiveness in addressing JTC challenges.