关键词: 目标跟踪, KCF, 增量式模板更新, 尺度自适应

Abstract: Aiming at the problem that the kernel correlation filtering target tracking algorithm (KCF) fails to track when the target scale changes, is severely occluded, and out of the field of view, the anti-jamming design is based on the kernel-related filtering tracking algorithm from two aspects. Aiming at the problem that KCF's template update method is likely to cause template pollution when the target is severely occluded, an incremental template update strategy is proposed to enhance the robustness of the algorithm; for KCF's lack of ability to deal with target scale changes, scale adaptive estimation is carried out. To adapt to changes in the target form. Finally, the feasibility and effectiveness of the two anti-jamming designs are verified and analyzed through experiments

Key words: Target tracking, KCF, Incremental template update, Scale adaptation