Research On Visual Pose Measurement For Adapter Ring

Space robotics plays a critical role in current and future space exploration missions.It improves humanity’s ability to explore and to operate by providing access beyond human limitations in the harsh environment and supporting operations that extend astronauts’ capabilities.One broad area of application is space construction,on-orbit maintenance and abandoned debris de-orbiting.All the above-mentioned applications justify the development of fast and robust pose measurement method during the capturing phase.By abstracting the adapter ring into a circular feature,this paper studies the visual system of the space robot,including both eye-to-hand and eye-in-hand cameras.The sunlight is intense and highly directional in outer space.Besides,the target satellite always covers with a sheet called Multi Layer Insulation(MLI)to insulate against thermal losses,leading to a strong reflection as well as a complex background.Moreover,the on-orbit computational resource is always limited,even much less than the mobile phones.Thus,this paper presents a fast and robust ellipse detector based on edge following method.It first extracts edge segments automatically based on the2 D iamge entropy.The detection result is a set of edge segments rather than a binary edge map,each of which is a clean,contiguous,1-pixel wide chain of pixels.Then,line segments are generated based on the length condition instead of least squares errors.This step can reduce the number of edge points while keeping the desired accuracy.After that,potential ellipses are detected based on edge curvature and convexity.In addition,a re-find contours detection method is introduced to improve the accuracy by searching edge points in the missing part of the ellipse.A stereo vision system is presented to estimate the non-cooperative satellites based on the circular feature.It works as an eye-to-hand vision camera in the final approach phase.Given the intrinsic parameters as well as the ellipse equations,the position and orientation of the adapter ring can be determined directly in quadratic form.Neither additional sensors nor prior knowledge is required,and the orientationduality problem has been solved by means of a revised defination of Eular angles.To enhance the safety of OOS operations,this paper also discusses the pose measurement method when one of the binocular cameras breaks down.During proximity operations,this circle feature may be occluded by the robot arm or limited field of view.In this paper,a structured light vision system is proposed to estimate the pose of the non-cooperative satellites,which is comprised of a monocular camera and three laser projectors.These projectors are used to emit lines on the surface of the satellite.Six break points will appear along the inner/outer edges of the adapter ring.A closed-form solution is then provided using the least square method instead of traditional geometric intersections.A real-time virtual simulation platform is built up to simulate synthetic images of the target satellite.It adopts the computer-graphics technology to illustrate the relative relationship and takes advantage of cheap costs,reliable technology,flexibility and convenience.Furthermore,a ground experiment system is set up to verify the effectiveness and robustness of the above-mentioned methods.The visual system of the space robot is tested on this platform,including both eye-to-hand and eye-in-hand cameras.Experimental results indicate that it is a practical and effective method for the pose measurement of non-cooperative satellites.The detection result is smoothed by a kalman based filter to increase the tracking accuracy of the space robot.Moreover,the grey forecasting model is introduced to deal with the wrong/undetected results.