Satellite video target tracking based on kernel correlation filtering and trajectory prediction

There are various problems in the current target tracking of satellite videos, such as the small size of moving targets, few textures, and easy occlusion. To solve the problem of tracking failure caused by occlusion of moving objects, a tracking algorithm based on kernel correlation filtering and trajectory prediction is proposed in this paper. The algorithm in this paper uses the motion characteristics of moving objects in satellite video to predict the trajectory and uses the trajectory prediction algorithm to predict the trajectory of the target when it is occluded. Combined with the kernel correlation filtering algorithm, the target can still be accurately tracked after it is occluded. Experiments on satellite video datasets show that the algorithm in this paper can effectively solve the problem of tracking failure caused by the occlusion of moving objects, and the speed is also very fast.