Motion Parameter Estimation And 3D-reconstruction Of Non-cooperative Space Objects

With the rapid development of space technology and the increasing of space activities,such as satel ite on-orbit assembly and maintenance for the purpose of space objects on-orbit capture technology has become an important research direction in the field of space technology.The motion parameter estimation and 3Dreconstruction of non-cooperative space objects is one of the main key technologies in the field of orbit trapping technology,which has important research value and significance.The main research work of this paper is based on the image information of binocular stereo camera,and realizes the motion parameter estimation and 3Ddimensional reconstruction of non-cooperation space objects.Firstly,a 3D point cloud acquisition system based on binocular stereo camera is constructed.The camera imaging model is described and the camera calibration is completed;Scheme of Initial 3D point clouds was designed using the Triclops stereo vision library,through the preprocessing and the three-dimensional point cloud processing,obtain parallax information of the objects,and based on the triangulation method to calculate the objects depth information,and then get objects initial threedimensional point cloud;To study the 3D point cloud post-processing method based on the PCL,to complete the objects point cloud denoising,splicing and sampling.Secondly,the paper researches a kind of motion parameters estimation algorithm based UT Kalman filter.The cartesian position of objects geometric center and attitude of principal inertial of the point cloud can be obtained by processing the 3D point cloud.The two types of data are taken as the observation,combined with the objects on-orbit kinematics and dynamics model,to estimate the accurate and complete movement parameter of the space objects with UT Kalman filter.The numerical simulation results show that the proposed method has high robustness and accuracy for the estimation of motion parameters of non-cooperative space objects.Then,a method to improve the speed and accuracy of non-cooperative space objects 3D reconstruction is researched.Design a point cloud simplification algorithm based on quadratic grid,through the second grid to space division of point cloud,the k-nearest neighbor search algorithm to search for k-nearest neighbor of data points,and then calculate the normal vector information of point cloud,according to the angle between the normal vector to the selectivity of point cloud sampling,can preserve the geometric characteristics of the objects,final y finishing the objects surface reconstruction using Power Crust algorithm.Having carried on the simulation study aiming at typical nozzle in space objects characteristic,verify that the method can removing large amounts of redundant data in the point cloud,and realize the fast and accurate 3D reconstruction of the spatial objects point cloud at the same time.Finally,an experimental platform is proposed to validate the motion parameter estimation and 3D-reconstruction algorithm of non-cooperative space objects.Using the robot arm to capture the satel ite model to simulate its movement in space,through the objects image information collected at different views,final y obtain the objects motion parameters,and to get to the point cloud of satel ite model to simplify and 3D-reconstruction,reshape its geometric shape.The analysis of the experimental results shows that the proposed method has high robustness and accuracy for the estimation of the motion parameters of the non-cooperative space objects,and can realize the quickly and accurate 3D-reconstruction of the objects point cloud.