Research on Video Pedestrian Tracking Based on the Combination of Optical Flow Method and Target Tracking Network

Emily Zhao,Dazhi Zhang,Li Yao, Yihang Guo,Boying Wu,Zhichang Guo, Jie Ning

Lecture notes in networks and systems(2023)

引用 0|浏览2
暂无评分
摘要
Target tracking is designed to predict the position of a given target in each video frame, and it has a wide range of applications in the fields of robot vision, video surveillance, and unmanned driving. This paper mainly studies the challenges of interference and target occlusion of similarities in target tracking, takes pedestrians as the main research target, combines the optical flow algorithm with the target tracking network, and improves the optical flow algorithm and the target tracking model, thereby improving the robustness of the tracking algorithm. Extensive experiments and comparative tests show the effectiveness of our method.
更多
查看译文
关键词
video pedestrian tracking,optical flow method,target tracking network
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要