Tracking of Maneuvering Extended Object with Random Hypersurface Model Based on Adaptive Uncorrelated Conversion Filter

Maneuvering extended object tracking is a rapidly developing research field, owing to the advent of high-resolution sensors, and is of significant value in both civilian and military applications. However, current extended object tracking techniques are limited by the linear minimum mean square error estimation framework, which constrains their tracking performance. And the unknown time-varying characteristics of the system noise will also bring the problem of imprecise estimation for the tracking. Thus, this study proposes a novel approach to track maneuvering extended objects using a star-convex random hypersurface model based on an adaptive uncorrelated conversion filter. The proposed method employs the additional measurement information, thereby breaking the linear estimator framework, and an adaptive strategy is also introduced to solve the noise interference problem. The simulation results demonstrate the efficacy of the proposed approach, which enhances the accuracy of the object’s kinematic state and extension estimation, compared to several commonly used filters.