Extended target tracking using star-convex model with nonlinear inequality constraints

Due to the increased resolution capability of modern sensors, the assumption that the received measurement originated from a point source may be invalid. Thus extended target tracking becomes more and more important in many practical scenarios. A novel extended target model has been proposed that describes the target approximately using a star-convex shape. Although the star-convex model can model a wide spectrum of shapes, the estimation results will be unpredictable without imposing constraints. This paper attempts to solve this problem by introducing a modified star-convex model with two proposed constraints. Simulation results demonstrate the effectiveness of the proposed tracking algorithms.