Multi-scale Correlation Filter Tracking Method Based On Occlusion Discrimination

A novel approach to predictive tracking,which is based on occlusion discriminant multi-scale correlation filter tracking method,is proposed to overcome the problems of low accuracy caused by occlusion,which are scale changes in the tracking process in complex environments.First,the foreground area of the first frame image is selected,and the target position and scale filters and GMS detector are trained.Then,the target location is estimated by the translation filter,and the target scale is calculated by the scale filter.Performing a correlation between the candidate samples that are obtained using different scales center on the new position,and the scale correlation filter derives the primary target area.Finally,the primary target area is evaluated by the correlation filter response,and occlusion and update conditions are determined by the peak and multiple peak fluctuation of the correlation filter response values.If the target need update,then this update should focus on the location and scale filters and the GMS detector to complete the target model update.If the target is occluded,it is necessary to further judge the occlusion degree of the target.In the case of partial occlusion of the target,this paper adopts the correlation filter tracking method to track the target.In the case of serious occlusion and total occlusion of the target,the detector detects the target position to complete target location.Experiments show that the method has a lower average center error of 5.47 and has a better tracking accuracy of 0.916 and tracking speed of 26.2 frames per second,compared with six state-of-the-art tracking algorithms,the method shows improved tracking performance with higher tracking speed and accuracy.It sloves the target loss problem in the occlusion and the problem of model updating errors.The method can track the target rapidly and accurately under conditions of varying scale transformation and occlusion,and it has enhanced tracking accuracy and robustness.There are 24 figures,8 tables and 67 references in this paper.