中文摘要：

提出了一种利用绕飞段拍摄序列图像对小天体三维模型进行重构,同时估计探测器位置姿态的方法。利用PCA-SIFT算法对探测器绕飞小天体过程中拍摄的序列图像进行特征点提取及匹配,得到稳定的特征点序列。在窄视场拍摄条件下,通过弱透视投影变换简化成像模型,并利用SVD因式分解算法重构了小天体的三维模型,同时估计出探测器的位置和姿态。通过实际拍摄的图像以及人工数据分析了距离远近、特征点个数、图像帧数以及视场角等因素对三维重构精度以及探测器位姿估计的影响。仿真结果表明该算法可以有效的用于小天体模型重构以及探测器的运动估计。

英文摘要：

In order to reconstruct a 3D small body model and estimate the spacecraft pose during fly-around the small body,a method based on an image sequence is proposed.Scale-invariant feature points from the image sequence are extracted and matched by PCA-SIFT algorithm.As the images are taken by narrow angle camera,image formation model can be simplified by a weak perspective projection transform,and then the 3D small body model and the spacecraft position and attitude can be recovered by using SVD factorization algorithm.Real images and synthetic data are used to analyze recovery error affected by distance,number of feature points,image frames and field angle etc.Simulation results show that this method can be used for reconstruction of the 3D small body model and the spacecraft motion estimation effectively.