Open AccessResearch on Depth Image Tracking Based on Aerial Refueling
Author(s) -
Shuyao Zhang,
Xianyun Qian,
Jun Wang,
Fan Zhang
Publication year - 2022
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2246/1/012080
Subject(s) - computer vision , artificial intelligence , computer science , robustness (evolution) , grayscale , kalman filter , tracking (education) , image (mathematics) , psychology , pedagogy , biochemistry , chemistry , gene
Aiming at the problem of the current aerial refueling docking which is difficult to accurately identify the floating cone sleeve in a complex environment, the Meanshift algorithm can efficiently track moving targets in the plane, there is not very good solution strategy for the pose measurement of the cone sleeve and the movement in the optical axis direction. The Time-Of-Flight (TOF) in the depth camera can directly collect depth images and grayscale images, providing the distance information displayed by a flat camera. By using the Meanshift algorithm to iterate in the three-dimensional space, the Kalman filter is added to achieve robustness to the tracking of moving targets, effectively solve the problem of occlusion of the target, and provide the coordinate position of the target. In the experiment, 600 frames of images are extracted for comparison. The effect of the TOF camera has higher accuracy and stability than that of the flat camera. It still has a stable tracking effect after using the props to cover the cone model, which is better than the traditional plane Meanshift algorithm.