基于改进的HD距离的星图识别算法

Star Chart Recognition Algorithm Based on Improved Hausdorff Distance

在基于星敏感器的航天飞行器的自主天文导航中,星图识别是关键技术之一。将改进的HD距离算法用于星图识别,避免了由于个别星体出现较大噪声,以及流星等干扰星出现时发生误匹配的问题。在匹配时利用了星图的全部空间结构信息,具有强的抗星敏感器畸变、噪声和旋转能力,同时也不需要复杂的特征提取算法。仿真实验显示它具有较高的实用价值。

Among the methods of automatical celestial navigation which works on the spacecraft based on the star sensor, the star chart recognition is one of the key technologies. The improved Hausdorff distance algorithm used for identification can avoid the problem of mismatch while information of individual star was affected by noise and a few meteors exist. This method has taken into account for information of star dimensional configuration in matching. Therefore, the approach has strong robustness for the disturbance of noise, distortion, and rotation of lens of the star sensor. Moreover, it does not need any complex algorithm for feature extraction. The simulation results show that itis of high value of practice.