Aiming at the difficult real-time tracking and low positioning accuracy of mobile targets on the sea,in order to further broaden the supervision of offshore areas and ensure the safety of ad-jacent seas,the optimal design and implementation of ship radar surveillance systems are worth in-tensive research.Through the research into architecture analysis,functional design,etc.of a ship radar surveillance system,multi-source data fusion technology has been utilized in this paper to a-chieve real-time,dynamic and three-dimensional block-based intelligent monitoring of offshore ship targets.Aiming at the core issue of ship maneuvering target tracking,a multi-target track initiation algorithm based on nonlinear unsented Kalman Filter(UKF)filtering is proposed in this paper,which can effectively improve the probability of detecting targets and reduce the false tracks.The effectiveness of the algorithm is verified by simulation.This study has certain reference value.