摘要
近十年来,深度学习由于其优越的性能,已经逐渐应用于各个领域。在目标跟踪领域,基于深度学习的方法也取得了巨大的成功。文章主要介绍基于深度学习的目标跟踪算法研究现状及发展趋势。首先,介绍了视觉目标跟踪传统算法。然后,对基于深度学习的目标跟踪算法进行分类,并进行问题分析。最后,对基于深度学习的目标跟踪算法的发展趋势进行预测。
In recent ten years,deep learning has been gradually applied in various fields because of its superior performance.In the field of target tracking,the method based on deep learning has also achieved great success.This paper mainly introduces the research status and development trend of target tracking algorithm based on deep learning.Firstly,the traditional algorithms for visual target tracking are introduced.Then,the target tracking algorithms based on deep learning are classified and the problems are analyzed.Finally,the development trend of target tracking algorithm based on deep learning is predicted.
作者
张莹杰
ZHANG Yingjie(Zhanjiang University of Science and Technology,Zhanjiang 524094,China)
出处
《现代信息科技》
2021年第8期82-85,共4页
Modern Information Technology
关键词
目标跟踪
深度学习
孪生网络
相关滤波
target tracking
deep learning
siamese network
correlation filtering
