智能无杆飞机牵引车视觉定位算法

为了在智能无杆飞机牵引车中利用单目视觉实时检测牵引车相对飞机前轮的位置信息，实现无杆牵引车与飞机的自动对接，在研究基本几何理论的基础上，建立了一种飞机轮胎的近似几何模型。通过对模型的分析，提出利用飞机轮胎平面图像检测车身相对机轮偏角的算法，以及利用像平面与地平面的坐标映射关系检测车身中线相对机轮偏距的算法。算法仿真结果显示，该基于单目视觉的定位算法具有较高的准确性，可以为系统提供满足精度要求的横向控制参数。

Vision Localization Algorithm of Intelligent Towbarless Aircraft Tractor

This paper dedicates to use the intelligent aircraft tractor’s monocular vision to detect the real-time relative position of the tractor and aircraft’s front wheel, and to realize the towbarless tractor auto connects the aircraft’s front wheel. On the basis of researching basic geometry theory, this paper establishes an approximate geometry model of the aircraft’s tire. Through analysis of the approximate geometry model, it presents an algorithm which uses the image of aircraft’s tire to detect the relative angle of the tractor and aircraft’s front wheel. Using the mapping relation of the image plane and ground plane, it presents an algorithm to detect the setover between the tractor’s midline and aircraft’s front wheel. And the simulation result proves that the localization algorithm can get much more accurate data and provide satisfied lateral control parameters for the intelligent towbarless aircraft tractor.