Multitarget tracking algorithm based on LDL decomposition

Random target number, heavy clutter, and heavy target density make difficult for multiple targets tracking. Gaussian mixture Cardinalized probability hypothesis density (GMCPHD) filter based on theory of finite set statistics can perform this tracking task while with complex computation. It is well known that the computation of square-root Kalman filter based on LDL decomposition is about the half of the one in extend Kalman filter (EKF). Hence, this paper design a multitarget tracking algorithm based on LDL decomposition which called LDL-GM-CPHD filter. Numerical example is given in this manuscript and the simulation results reveal that LDL-GM-CPHD filter can save about 30% time resource with similar tracking performance.