Abstract: In the process of linear array scanning laser radar reconstruction for the 3D model of non-cooperative unstable spacecraft, there is a problem that the moving target line scan data is distorted and it is difficult to accurately estimate the target’s motion characteristics due to instability. Therefore, distortion tolerance analysis is needed and a self-correcting three-dimensional reconstruction method with distortion is proposed. The method is based on the angular precision and ranging accuracy of the measuring condition to design the distortion tolerance threshold value judgment system. Firstly, according to the actual working conditions and the judgment condition, it is determined whether the initial distortion recovery is performed on the line scan data; secondly, the adjacent scanning band distortion point cloud is sequentially registered to obtain the target motion increment and solve the target instantaneous motion parameter; then, the target motion parameters are self-corrected to remove the characteristic mutation point interference; at the same time, the point cloud is sequentially restored to the initial geometric position, and the initial target three-dimensional reconstruction is completed; finally, through further iteration of the system, the pose increment self-converges until it finally meets the reconstruction accuracy requirements. The simulation results show that the proposed method can effectively avoid the distortion caused by line scan in the absence of non-cooperative target prior information, and achieve the target 3D reconstruction that meets the accuracy requirements quickly, which has strong robustness.

Key words: 3D reconstruction, Line array imaging, Self-correction with distortion, Distortion tolerance