Research On Target Tracking Method Based On Satellite Video

With the development of remote sensing technology,the application field of remote sensing products has beeing expanded,and the remote sensing products of image form have been unable to meet the needs of all walks of life.As target tracking is an important research field of video application,so the research on target tracking method based on satellite video has important theoretical significance and application value.Aiming at the problem that single target in satellite video which has smaller scale is difficult to track,according to the chromaticity difference weighted idea,we put forward the improved CAMSHIFT target tracking algorithm,and verified the algorithm for satellite video target tracking is feas’ible and the tracking effect is better than the traditional CAMSHIFT algorithm through experiments.For the problem that multpile target tracking in satellite video is easily affected by noise,according to the idea of filtering noise,improved mixture Gaussian background model algorithm is proposed.And we verified that this algorithm is applicability and efficient using in satellite video target tracking,and has better tracking performance.At the end of this paper,supported with the improved CAMSHIFT target tracking algorithm and the improved Gaussian mixture background model algorithm,with visual studio as the development tool,combined with the opencv computer vision library,under the windows platform,we designed and developed the satellite video target tracking system which can be used for a wide range of traffic monitoring.