海上星光导航是航海中一种重要的自主导航技术，星图识别是其关键步骤。针对船载鱼眼相机星光导航系统超大视场带来的单幅图像数据量大、识别冗余、识别效率低等问题，提出了一种基于圆形视场分割的鱼眼相机星图识别方法。对于拍摄到的星图，利用同心圆将视场分割成若干个面积相等的环形和圆形区域；在构造导航星特征库的过程中，以星角距为特征构造散列函数，将导航特征库分段存储成若干个子库；在识别过程中，利用基于中心星的多三角形识别算法，从视场中心圆形区域开始依次向视场边缘环形区域进行识别。海上观测实验结果表明：该方法能够平均以2.5 s的识别时间达到90%以上的识别成功率，且具有良好的实时性。

Maritime celestial navigation is an important autonomous navigation technology at sea, and the star identification is a key step in celestial navigation. Aiming at the problems of large data, redundant identification and low speed caused by the large field of view (FOV) of the fisheye camera celestial navigation system, a fisheye camera star identification method based on circular FOV segmentation was proposed. For the star image taken by a fisheye camera,the concentric circles were drawn around the center with different radii, and the FOV was divided into several equal-area circular areas. In the process of constructing the navigation feature database, the star argument was taken as the feature to construct a hash function, and the navigation feature database was stored into several sub-databases. In the identification process, a multi-triangulation identification algorithm based on the central star was used to proceed from the center to the edge of the circular FOV. The results of marine observation experiments show that the method can achieve an identification success rate of more than 90% with the time of 2.5 s on average, and it has good real-time performance.