视频目标跟踪综述
张丰,冯平
摘 要: 研究近年来主流的目标跟踪算法。通过文献阅读和归纳对比,分析了使用生成式模型和判别式模型的目标跟踪算法。结果显示,对于存在复杂干扰因素的场景,采用第二类模型的目标跟踪算法的跟踪效果更好。文章为视频跟踪领域的研究者们提供了一个关于目标跟踪算法的客观分析。
关键词: 目标跟踪; 深度学习; 相关滤波; 计算机视觉
中图分类号:TP399 文献标识码:A 文章编号:1006-8228(2022)01-32-04
Overview on video target tracking
Zhang Feng, Feng Ping
(Guizhou University of Finance and Economics, Guiyang, Guizhou 550025, China)
Abstract: This paper focuses on the mainstream target tracking algorithms in recent years, and analyzes the target tracking algorithms using generative and discriminative models through literature reading and inductive comparison. The results show that for the scenes with complex in ……此处隐藏7423个字…… >
[13] Nam H,Han B. Learning Multi-Domain ConvolutionalNeural Networks for Visual Tracking[J],2015
[14] Song Y B, Ma C, Wu X H, Gong L J, Bao L C, Zuo W M,et.al. Visual tracking via adversarial learning. In: Proceedings of the 2018 IEEE Conference on Computer Vision and Pat-tern Recognition. Salt Lake City, Utah, USA:IEEE,2018:1084−1093
[15] Bertinetto L, Valmadre J, Henriques J F, et al. Fully-convolutional siamese networks for object tracking[C]. European conference on computer vision,2016:850-865
[16] Li B, Yan J, Wu W, et al. High performance visual tracking with siamese region proposal network[C].Proceedings of the IEEE Conference of Computer Vision and Pattern Recognition,2018