无人机空中对接中的视觉导航方法
|
摘要
无人机空中加油是一种能有效提升巡航里程及续航时长的技术手段,近距相对位置和姿态测量技术是其中需要解决的关键问题之一。针对该问题研究了无人机自主空中对接中的视觉导航方法,完成了近距对接的地面实验。首先,利用移动对接图标和无人机的GPS/INS信息进行无人机的粗略导航,完成会合;再充分利用视觉图标的颜色与形状信息,通过颜色分割选取目标可能区域,在这些区域中进行快速椭圆检测,针对椭圆检测算法存在误检测及边缘不重合的问题,提出了椭圆检测与轮廓检测相结合的方法,能够更准确地描述图标边缘;最后,利用改进的OI算法进行相对位姿的估计,实现近距的精确导航。实验结果表明,无人机在较高速度下的跟踪效果良好,采用的视觉导航方法能够满足空中对接中精度与实时性的要求。
UAV air refueling is a technical means to effectively improve the range and duration of the cruising.Close-range relative position and attitude measurement technology is one of the key issues that need to be solved.In this paper,a visual navigation method in the autonomous air docking of drones is studied,and the ground experiment of close docking is performed.Firstly,the mobile docking icon and the GPS/INS information of the drone are used for rough navigation of the drone to complete the meeting;then the color and shape of the visual icon is fully utilized,target possible areas are selected by color segmentation.In these areas,fast ellipse detection is carried out.In ellipse detection algorithm,there are problems of false detection and edge misalignment.A method combining ellipse detection and contour detection is proposed.Finally,an improved OI algorithm is used to estimate the relative pose and attitude.The experimental results show that the tracking effect of the UAV is good at higher speeds,and the visual navigation method can meet the requirements of accuracy and real-time in the drone air docking.
UAV air refueling is a technical means to effectively improve the range and duration of the cruising.Close-range relative position and attitude measurement technology is one of the key issues that need to be solved.In this paper,a visual navigation method in the autonomous air docking of drones is studied,and the ground experiment of close docking is performed.Firstly,the mobile docking icon and the GPS/INS information of the drone are used for rough navigation of the drone to complete the meeting;then the color and shape of the visual icon is fully utilized,target possible areas are selected by color segmentation.In these areas,fast ellipse detection is carried out.In ellipse detection algorithm,there are problems of false detection and edge misalignment.A method combining ellipse detection and contour detection is proposed.Finally,an improved OI algorithm is used to estimate the relative pose and attitude.The experimental results show that the tracking effect of the UAV is good at higher speeds,and the visual navigation method can meet the requirements of accuracy and real-time in the drone air docking.
引文
[1]李怡勇,沈怀荣,高飞.发展高空长航时无人机初探[J].飞航导弹,2005(8):16-20.Li Yiyong,Shen Huairong,Gao Fei.Development of high-altitude and long-endurance drones[J].Winged Missiles Journal,2005(8):16-20(in Chinese).
[2]李大伟,王宏伦.无人机自动空中加油飞行控制技术[J].系统仿真学报,2010,22(S1):126-130.Li Dawei,Wang Honglun.UAV Flight control in automated aerial refueling[J].Journal of System Simulation,2010,22(S1):126-130(in Chinese).
[3]Nalepka J,Hinchman J.Automated aerial refueling:extending the effectiveness of UAVs[C]//AIAAModeling and Simulation Technologies Conference and Exhibit.San Francisco,2005:6005-6013.
[4]Hinchman J,Schreiter D.Automated aerial refueling[C]//2007Aerial Refueling Systems Advisory Group Conference.Las Vegas:Air Force Research Laboratory,2007:1-18.
[5]Wang X,Kong X,Zhi J,et al.Real-time drogue recognition and 3Dlocating for UAV autonomous aerial refueling based on monocular machine vision[J].Chinese Journal of Aeronautics,2015,28(6):1667-1675.
[6]Xu Y,Duan H,Li C,et al.On-board visual navigation system for unmanned aerial vehicles autonomous aerial refueling[C]//Proceedings of the Institution of Mechanical Engineers,Part G:Journal of Aerospace Engineering,2017.
[7]Chen S,Duan H,Deng Y,et al.Drogue pose estimation for unmanned aerial vehicle autonomous aerial refueling system based on infrared vision sensor[J].Optical Engineering,2017,56(12):124105.
[8]纪超,王庆.基于双目视觉的自主空中加油算法研究与仿真[J].系统仿真学报,2013,25(6):1327-1331.Ji Chao,Wang Qing.Stereo vision system and simulation for autonomous aerial refueling[J].Journal of System Simulation,2013,25(6):1327-1331(in Chinese).
[9]李柱.基于双目视觉的无人机自主空中加油对接导航方法[D].厦门:厦门大学,2017.Li Zhu.Binocular-vision-based docking navigation method for UAV self-refueling[D].Xiamen:Xiamen University,2017(in Chinese).
[10]周婷.基于双目视觉的无人机空中加油特征匹配方法研究[D].厦门:厦门大学,2014Zhou Ting.Feature matching research of the autonomous aerial refueling based on binocular vision[D].Xiamen:Xiamen University,2014(in Chinese).
[11]Duan H,Li H,Luo Q,et al.A binocular vision-based UAVs autonomous aerial refueling platform[J].Science China Information Sciences,2016,59(5):053201.
[12]Li H,Duan H B.Verification of monocular and binocular pose estimation algorithms in vision-based UA-Vs autonomous aerial refueling system[J].Science China Technological Sciences,2016,59(11):1730-1738.
[13]Sun S,Yin Y,Wang X,et al.Robust landmark detection and position measurement based on monocular vision for autonomous aerial refueling of UAVs[J].IEEE Transactions on Cybernetics,2018(16):1-13.
[14]Fornaciari M,Prati A,Cucchiara R.A fast and effective ellipse detector for embedded vision applications[J].Pattern Recognition,2014,47(11):3693-3708.
[15]周城宇,甄子洋,黄一敏,等.基于视觉引导的无人直升机着舰技术研究[J].导航定位与授时,2018,5(4):43-48.Zhou Chengyu,Zhen Ziyang,Huang Yimin,et al.Research on landing technology of unmanned helicopter based on vision guidance[J].Navigation Positioning&Timing,2018,5(4):43-48(in Chinese).
[16]刘崇亮,刘垒,扈光锋,等.一种着陆视觉导航P3P问题的解法[J].导航定位与授时,2018,5(1):64-67.Liu Chongliang,Liu Lei,Hu Guangfeng,et al.AP3Pproblem solving algorithm for landing vision navigation[J].Navigation Positioning&Timing,2018,5(1):64-67(in Chinese).
[17]Lu C P,Hager G D,Mjolsness E.Fast and globally convergent pose estimation from video images[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2000,22(6):610-622.
[18]Schweighofer G,Pinz A.Robust pose estimation from a planar target[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(12):2024-2030.
[2]李大伟,王宏伦.无人机自动空中加油飞行控制技术[J].系统仿真学报,2010,22(S1):126-130.Li Dawei,Wang Honglun.UAV Flight control in automated aerial refueling[J].Journal of System Simulation,2010,22(S1):126-130(in Chinese).
[3]Nalepka J,Hinchman J.Automated aerial refueling:extending the effectiveness of UAVs[C]//AIAAModeling and Simulation Technologies Conference and Exhibit.San Francisco,2005:6005-6013.
[4]Hinchman J,Schreiter D.Automated aerial refueling[C]//2007Aerial Refueling Systems Advisory Group Conference.Las Vegas:Air Force Research Laboratory,2007:1-18.
[5]Wang X,Kong X,Zhi J,et al.Real-time drogue recognition and 3Dlocating for UAV autonomous aerial refueling based on monocular machine vision[J].Chinese Journal of Aeronautics,2015,28(6):1667-1675.
[6]Xu Y,Duan H,Li C,et al.On-board visual navigation system for unmanned aerial vehicles autonomous aerial refueling[C]//Proceedings of the Institution of Mechanical Engineers,Part G:Journal of Aerospace Engineering,2017.
[7]Chen S,Duan H,Deng Y,et al.Drogue pose estimation for unmanned aerial vehicle autonomous aerial refueling system based on infrared vision sensor[J].Optical Engineering,2017,56(12):124105.
[8]纪超,王庆.基于双目视觉的自主空中加油算法研究与仿真[J].系统仿真学报,2013,25(6):1327-1331.Ji Chao,Wang Qing.Stereo vision system and simulation for autonomous aerial refueling[J].Journal of System Simulation,2013,25(6):1327-1331(in Chinese).
[9]李柱.基于双目视觉的无人机自主空中加油对接导航方法[D].厦门:厦门大学,2017.Li Zhu.Binocular-vision-based docking navigation method for UAV self-refueling[D].Xiamen:Xiamen University,2017(in Chinese).
[10]周婷.基于双目视觉的无人机空中加油特征匹配方法研究[D].厦门:厦门大学,2014Zhou Ting.Feature matching research of the autonomous aerial refueling based on binocular vision[D].Xiamen:Xiamen University,2014(in Chinese).
[11]Duan H,Li H,Luo Q,et al.A binocular vision-based UAVs autonomous aerial refueling platform[J].Science China Information Sciences,2016,59(5):053201.
[12]Li H,Duan H B.Verification of monocular and binocular pose estimation algorithms in vision-based UA-Vs autonomous aerial refueling system[J].Science China Technological Sciences,2016,59(11):1730-1738.
[13]Sun S,Yin Y,Wang X,et al.Robust landmark detection and position measurement based on monocular vision for autonomous aerial refueling of UAVs[J].IEEE Transactions on Cybernetics,2018(16):1-13.
[14]Fornaciari M,Prati A,Cucchiara R.A fast and effective ellipse detector for embedded vision applications[J].Pattern Recognition,2014,47(11):3693-3708.
[15]周城宇,甄子洋,黄一敏,等.基于视觉引导的无人直升机着舰技术研究[J].导航定位与授时,2018,5(4):43-48.Zhou Chengyu,Zhen Ziyang,Huang Yimin,et al.Research on landing technology of unmanned helicopter based on vision guidance[J].Navigation Positioning&Timing,2018,5(4):43-48(in Chinese).
[16]刘崇亮,刘垒,扈光锋,等.一种着陆视觉导航P3P问题的解法[J].导航定位与授时,2018,5(1):64-67.Liu Chongliang,Liu Lei,Hu Guangfeng,et al.AP3Pproblem solving algorithm for landing vision navigation[J].Navigation Positioning&Timing,2018,5(1):64-67(in Chinese).
[17]Lu C P,Hager G D,Mjolsness E.Fast and globally convergent pose estimation from video images[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2000,22(6):610-622.
[18]Schweighofer G,Pinz A.Robust pose estimation from a planar target[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(12):2024-2030.