计算机科学
跟踪(教育)
计算机视觉
人工智能
帧(网络)
理论(学习稳定性)
帧速率
弹道
磁道(磁盘驱动器)
跟踪系统
跑道
卡尔曼滤波器
电信
历史
操作系统
教育学
机器学习
物理
心理学
考古
天文
作者
Hui Yao,Dong He,Yunfeng Sui,Shixuan Zhao
标识
DOI:10.1109/iccasit55263.2022.9986851
摘要
For the visual targets in the runway and taxiway, we designed a tracking system, which can effectively track the targets stably and smoothly, and ensure that the targets are always located in the central area of the picture. This paper will introduce the three main parts of the system in detail: target centralization, tracking algorithm and control strategy. Firstly, we proposed to quickly centralize the target before tracking, which can improve the stability. We also realized a fast centralization method with small deviation from the imaging principle of the camera. After that, we comprehensively balanced the advantages and disadvantages of classical tracking algorithm, selected the suitable one, and modified it to realize a method with a frame rate of 170 frames per second. In terms of control, we focused on the strategy of the direction and speed, which ensure the smooth movement of the running camera. Finally, we put forward some points which can be improved.
科研通智能强力驱动
Strongly Powered by AbleSci AI