一种面向空间飞行器视觉导航的椭圆检测算法

基于椭圆特征的空间飞行器视觉导航技术是一种新颖的高精度空间探测自主导航方法，如何对空间目标的环形边缘进行精准提取和高效拟合是实现空间飞行器视觉导航的必要条件。针对该问题，提出一种面向空间飞行器视觉导航的椭圆检测算法。利用多项式逼近导航图像连续边缘段的方式提取椭圆弧段；通过基于极大似然假设检验理论构建的模型选择判据，对来自同一个椭圆的椭圆弧段进行准确合并；对合并后的椭圆弧段进行拟合，得到空间飞行器视觉导航的椭圆检测结果。大量的仿真实验表明：与传统的椭圆检测算法相比，所提算法具有较高的精度和更高的鲁棒性，可以广泛应用于空间飞行器视觉导航图像椭圆检测，为空间飞行器视觉导航算法提供精准的二次曲线输入。

An ellipse detection algorithm for spacecraft optical navigation

The ellipse-based optical navigation technology has become a novel and precise autonomous navigation method. Therefore, how to fit the elliptic edge of the space object is the essential condition of the optical navigation method. We propose an ellipse detection algorithm for the spacecraft optical navigation in this paper. First, the elliptic curves are extracted by utilizing the polygonal curve to approximate the edge in edge map of the navigation image. Second, the elliptic curves are merged accurately by a model selection criterion derived from the maximal likelihood ratio hypothesis test. Finally, the ellipses in the navigation image are detected by fitting the merged elliptic curves. The experimental results demonstrate that the proposed algorithm can detect ellipses with higher accuracy and robustness than the traditional ellipse detection methods, and can be applied in the ellipse detection for the spacecraft optical navigation method to provide the precise curves as the inputs.