Algorithm of Ship Tracking based on the the filtering theory of Video Sequences

摘要

In order to realize real-time monitor of ship traffic situation to improve shipping safety and efficiency, reduce the burden on the driver and supervisor, this paper presents the Algorithm of ship tracking based on the the filtering theory of video sequences. The algorithm applies particle filter and mean shift algorithm to the technology of automatic ship tracking, at the same time it introduced the advantages and disadvantages between the Mean shift algorithm and particle filter algorithm by comparison. Experimental results show that the accuracy of particle filter tracking algorithm is lower than the mean shift algorithm, but it has better real-time. And also found that with the increase of the particles' number, the tracking accuracy is improved, but the running time becomes longer and the real-time gets worse.