An Improved Fast All-Sky Star Identification Based on the Polygon Algorithm

The star identification is a prerequisite for the in-orbit calibration of some space cameras. This paper proposes an improved polygonal fast all-sky star identification method with the unstable internal and external states of the space infrared camera in the early stage of orbit. By utilizing a searching strategy based on star-number lookup table combined with sorting, the conventional triple loop is transformed into a double small loop to accelerate the matching search process. At the same time, the constraint of polygonal edge is introduced to reduce the matching candidate set until it is unique, which ensures the successful recognition under the loose threshold of star angular distance. The experiment effectively proves that the recognition speed is significantly improved compared with the conventional method, which is only 1/5 of the original search time, less than 10s for the triangle matching of the dense star catalog. Meanwhile, the success rate of this method is still very high even when the stars are densely distributed and the camera internal geometric error is large.