In order to solve the problem that the crossover of target trajectory would make the algorithm more complex and make the tracing accuracy lower in the process of multitarget tracking. This paper proposed a crossmultitarget tracking algorithm based on random finite sets. The proposed algorithm substituted the Gaussian mixture probability parameter for the probability hypothesis density in order to decrease the complexity and educe the tracing error. The MATLAB simulation is implemented under linear Gaussian assumptions, this paper simulated the moving process of multiple targets, and used the improved algorithm proposed in this paper and the traditional data association algorithm for tracking multiple targets, the experimental results show that the proposed algorithm has better performance for the crossmultitarget tracking.