Research On Space Target Attitude Estimation And Stable Tracking Method

Relative attitude measurement and stable tracking of close-rotating space targets are difficult problems in autonomous rendezvous and docking of space non-cooperative targets.Affected by attitude movement,illumination changes,shape deformation and even noise,the fluctuations of characteristic points,characteristic lines and characteristic areas of non-cooperative rotating space targets such as satellites are obvious and their stability is weak.This leads to the traditional target tracking method often causing tracking point jitter,drift or even divergence,which is difficult to meet the high-precision tracking required by tasks such as close-range inspection,rendezvous and docking,and even capture.Therefore,the stable tracking of close-rotating space targets under uneven lighting conditions is bound to face many challenges such as changes in rotation,shape,posture,and motion blur,and has certain theoretical significance and practical value.Firstly,for the problem of relative attitude measurement of close-rotating space targets,this paper proposes a method to extract the target elliptical cross-sectional features,combined with the extended Kalman filter(EKF)and SLAM algorithm measurement methods to estimate the attitude;then for close-range rotating space targets In this paper,we use the method of extracting Harris corner features and extended Kalman filter(EKF)to achieve stable tracking of the target,and achieve high-stability real-time target tracking performance during the tracking process.This paper mainly covers the two research focuses of target attitude estimation and target stability tracking:(1)In the work of target pose estimation,this paper firstly extracts the feature region,feature line,feature point and other features of the rotating space target by extracting the spatial feature state of the close-space rotating space target model,by extracting SIFT and SURF Feature,combined with SLAM algorithm to estimate the rotation angle of the target;then a new feature of elliptical cross-section is proposed,which uses the contour to establish the target expansion state and the nonlinear characteristics of the extended Kalman filter(EKF)to estimate the attitude of the target.Through comparison,it is found that the new feature proposed in this paper has smaller error and higher robustness.(2)In the work of target stable tracking,this paper studies the method of merging Harris corner feature and extended Kalman filter(EKF)to track the target stably.In the experiment,the three different areas of global area,local area and corner area were tracked respectively.By comparing the tracking results of extracting the centroid feature and Harris corner feature of global area and local area,it is found that the tracking method of this paper has a pixel error Smaller and more stable tracking.The experiment in this paper uses MATLAB software and OPENCV visual library to write the algorithm code and its simulation work.The experimental data meets the design requirements of the algorithm.