ABSTRACT Aiming at the extended target tracking problem with irregular shapes, an extended target tracking algorithm that fuses random hypersurface model (RHM) and random matrix model (RMM) is proposed in the case where the priori shape of the target is unknown. First, the algorithm adopts the RMM method to track the extended target and then takes the reference points on the elliptical shape contour and constructs an ordered sequence. Second, the discrete Fourier transform (DFT) is performed on the sequence to find the coefficients of the harmonic components to obtain the elliptical shape parameters. Then, the ellipsoidal shape parameter is used as the a priori shape parameter of the star‐convex RHM. Finally, the effectiveness of the proposed algorithm is verified by simulation experiments. The proposed extended target fusion method provides a more accurate means for precise tracking and state prediction of extended targets.
Qi et al. (Wed,) studied this question.