Extended object tracking based on support functions and extended Gaussian images

This paper considers tracking of extended objects using down-range and cross-range extent measurements. For extended objects in radar or sonar tracking, existing elliptical modeling and rectangular modeling usually assume that the major axis of the object is parallel to its velocity vector. However, this may not be true in many practical applications. In view of this, we attempt to solve this problem by proposing two modeling approaches based on support functions and extended Gaussian images, respectively. The two approaches differ mainly in shape parametric representation for different objects and can be easily integrated into the extended object tracking framework which enables estimation of the kinematic state and object extension jointly. The effectiveness of the proposed modeling and estimation is illustrated by simulation results.