- 期刊
A Survey of Target Tracking Algorithms based on Correlation Filtering
摘要
Visual object tracking plays an important role in many computer vision applications. Discriminative target tracking method based on correlation filtering(CF) theory has become a research hotspot in the field of target tracking due to its robustness and efficiency. Discriminative correlation filter greatly improves tracking robustness by introducing feature representation, nonlinear kernel, scale estimation, spatio-temporal regularization and continuous convolution. This paper first introduces the basic CF theory and the basic framework of target tracking. Secondly, CF-based trackers are summarized by category. Thirdly, using the target tracking benchmark database (OTB-2013) video sequence to conduct algorithm comparison experiments, analyze and compare the performance of 9 typical different CF trackers in recent years. Finally, according to the current research status, it points out the possible future development trend of CF. Although target tracking based on correlation filter has been widely used in the field of tracking and has made some progress, target tracking is still a huge challenge due to the impact of complex scenes and dramatic changes in the appearance of the target itself. It is of great significance for the development of target tracking to study the correlation filtering tracking algorithm with high efficiency and robustness.