计算机工程 ›› 2013, Vol. 39 ›› Issue (7): 237-241.doi: 10.3969/j.issn.1000-3428.2013.07.053

• 人工智能及识别技术 • 上一篇    下一篇

基于红外探测技术的无人机视觉引导助降系统

杜 晶1a,1b,2,雷志辉1a,周 翔1a   

  1. (1. 国防科学技术大学 a. 图像测量与视觉导航湖南省重点实验室;b. 计算机学院,长沙 410073; 2. 光电控制重点实验室,河南 洛阳 471009)
  • 收稿日期:2012-03-07 出版日期:2013-07-15 发布日期:2013-07-12
  • 作者简介:杜 晶(1987-),男,硕士研究生,主研方向:机器视觉;雷志辉,副教授;周 翔,博士研究生
  • 基金项目:
    国家自然科学基金资助项目(60904084)

Unmanned Aerial Vehicle Vision Guidance Assisted Landing System Based on Infrared Detection Technique

DU Jing 1a,1b,2, LEI Zhi-hui 1a, ZHOU Xiang 1a   

  1. (1a. Hunan Key Laboratory for Image Measurement and Vision Navigation; 1b. College of Computer, National University of Defense Technology, Changsha 410073, China; 2. Science and Technology on Electron-optic Control Laboratory, Luoyang 471009, China)
  • Received:2012-03-07 Online:2013-07-15 Published:2013-07-12

摘要: 传统的助降导航系统无法满足无人机着舰挂锁对于精确性、抗干扰性及可重复性的要求。为解决该问题,提出一种基于红外探测技术的无人机视觉引导助降系统。系统硬件包括红外标志灯、高动态像机和红外滤光片,系统软件包括光团目标检测算法和双目交汇算法,其中,光团目标检测算法使用归一化负拉氏高斯算子实现光团目标的定位,双目交汇算法利用快速标定方案完成像机标定,采用双像机进行空间交汇。实验结果表明,该系统在距离理想着陆点200 m以内的测量精度小于5 cm,可满足助降要求。

关键词: 无人机, 视觉引导, 双目交汇, 红外标志灯, 测量精度

Abstract: Traditional guidance assisted landing systems have some problems, such as expensive price, low accuracy, poor anti-interference and so on, which can’t satisfy the requirement about accuracy, anti-jamming and repeatable landing on ship. In order to resolve this problem, this paper presents a Unmanned Aerial Vehicle(UAV) vision guidance assisted landing system based on infrared detection technique. System hardware includes infrared mark light, high-dynamic cameras and optical filters. System software includes light blobs detection algorithm and the binocular intersection algorithm. Light blobs detection algorithm uses Normalized Negative Laplacian of Gaussian(NNLOG) to detect and tracks centers of the infrared lamps in the images. The binocular intersection algorithm uses two cameras to find target, tracks target in synchronization and binocular fair. Shift calibrate method is used on project. Experimental results show that the measurement precision of this system is less than 5 cm in the last pivotal 200 m away from the ideal landing site. It can satisfy the assisted landing requirement.

Key words: Unmanned Aerial Vehicle(UAV), vision guidance, binocular intersection, infrared mark light, measurement accuracy

中图分类号: