Binocular-vision-based Docking Navigation Method For UAV Self-refueling

In recent years,UAV with its small size,low cost,good concealment and many other advantages quickly gained the favor of the world,and in the military and civilian areas are playing an increasingly important role,especially in the military,is widely used in intelligence gathering,enemy detection,and even precision strike and many other tasks.However,it is because of its small size,UAV maximum takeoff weight is severely limited,which makes it unable to carry enough fuel,so the UAV combat radius and life time are very limited.The birth of the UAV's self-refueling technology has effectively solved this problem by adding fuel to the UAV in the air through the air refueling aircraft,which greatly increases the life time and combat radius of the UAV and can be reduced the amount of fuel loaded when the UAV take-off,increase the amount of weapons and ammunition or detection equipment,greatly enhanced the UAV combat and detection capabilities.Close docking navigation is the most critical place for UAV's self-refueling technology,which requires obtain the pose data of fuel tubing taper sleeve with very high accuracy.In order to solve the problem of close docking navigation accuracy effectively,this paper presents a UAV's self-refueling docking navigation method based on binocular stereoscopic vision.The method first use the UAV's left and right cameras with filters to capture the image of fuel tubing taper sleeve with LED sign lights,and then the image is grayed out,binary segmentation,median filter and so on.And then use the connected area marking method based on regional growth to extract the centroid of LED sign lights imaging areas in image as a feature point.Then use the improved haar wavelet transform to describe the left and right image feature points,and stereo matching left and right image feature points.Finally,the binocular vision principle is used to obtain the three-dimensional coordinates of the LED sign lights in the world coordinate system,the least squares method is used to fit the space circle of the three-dimensional coordinates,and then the position and attitude of the fuel tubing taper sleeve taper are obtained.In this paper,the self-built experimental platform is used to simulate the scene of the UAV,s self-refueling close docking navigation based on binocular vision,and the experimental data are obtained show that the accuracy of the fuel tubing taper sleeve pose information obtained by this paper presents the close docking navigation method based on binocular vision meet the high precision requirement of the UAV's self-refueling close docking navigation.