Correlation filtering target tracking method based on context awareness and self-adaptive response

Abstract

The invention discloses a correlation filtering target tracking method based on context awareness and self-adaptive response. According to a target video sequence, under the condition that the initialstate of a target is given in a first frame, the state of the target is estimated in following video sequences; according to the method, first, for the t(th) frame, feature information is calculatedin the current frame according to a regional position of the previous frame; then, background information of a target region is calculated, a response target is calculated, and training classifiers are merged; next, according to the target position determined in the previous frame, the target position in the current frame is calculated; a filtering model is updated; and last, for all frames in thesequence, the steps 1 to 6 are repeated, and the target position in the (t+1)(th) frame is obtained. Through the method, context information of the target region is fully utilized to reduce boundaryeffects and self-adaptively change target responses, position drifts occurring due to a circulant matrix in sample data generation are reduced, and tracking drifts caused by occlusion, deformation andmotion blurring are effectively reduced.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a correlation filtering target tracking method based on context perception and adaptive response in the technical field of digital image target tracking. Background technique [0002] Object tracking has always been a classic research problem in the field of computer vision (Documents 1, 2), and is widely used in human-computer interaction, intelligent monitoring, visual navigation, precision guidance and other fields. With the widespread attention of scholars on object tracking technology, object tracking technology has made significant progress in the past few years, and scholars have proposed many excellent object tracking algorithms. However, affected by occlusion, illumination, fast motion, motion blur, rotation, deformation and other factors, there are still great challenges in the field of target tracking, and the robustness and accuracy of the algorith...

AI Technical Summary

Problems solved by technology

[0004] Although these algorithms have made great progress in tracking speed and tracking accuracy, the error samples caused by the boundary effect and the error samples obtained by cyclic shift during fast motion and occlusion will cause the classifier to discriminate not strong enough, resulting in tracking Drift, which affects the robustness of the algorithm

Images