Deep learning and geometric algorithm combined non-cooperative target relative pose estimation method

Abstract

The invention belongs to the field of spacecraft navigation, and particularly relates to a deep learning and geometric algorithm combined non-cooperative target relative pose estimation method, whichcomprises the steps: manually selecting 2D key points by utilizing a non-cooperative target data set, and reconstructing 3D model combination of a non-cooperative target by combining multi-view triangulation; re-projecting the 3D coordinates to an image plane to obtain 2D coordinate estimated values of the key points; designing a target detection and semantic segmentation network and a key point regression network, and performing supervised regression on 2D key point coordinates; and minimizing errors of 2D-3D corresponding point coordinates by using a nonlinear least square method to estimatea six-degree-of-freedom relative position and a relative attitude of the non-cooperative target. According to the method, deep learning and a geometric optimization algorithm are effectively combined, research of a high-precision and high-speed pose estimation algorithm based on a visual image can be realized, and the method is an innovative application of deep learning in the aerospace field.

Description

technical field [0001] The invention belongs to the field of spacecraft navigation, and particularly relates to a non-cooperative spacecraft relative pose estimation method combining deep learning and geometric algorithms. In combination, by designing a deep neural network, the research on the high-precision and fast pose estimation algorithm based on visual images is realized, and it assists in the capture of failed spacecraft and the maintenance of faulty satellites, etc. A number of space tasks. Background technique [0002] The problem of obtaining pose information of non-cooperative spacecraft in space has high research value in the fields of space confrontation, on-orbit maintenance, space assembly, and autonomous rendezvous and docking. The objective needs of natural security. On-orbit 6DOF attitude estimation of non-cooperative spacecraft, that is, relative position and relative attitude estimation, is an essential technical task for on-orbit servicing and space def...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the non-cooperative target's six-degree-of-freedom pose estimation, the traditional method is dependent on the iterative closest point (ICP) step, which is expensive and cannot meet the requirements of autonomy, as well as the traditional visual processing algorithm's dependence on manual features and fixed The matching program limits its performance in challenging environments such as lighting changes and complex model structures. The present invention provides a method for estimating the relative pose of a non-cooperative target based on deep learning and traditional geometric algorithms.

Images