Fast star extraction algorithm based on the local gray histogram

Star sensor is a common sensor in celestial navigation. Under complex working conditions, the star extraction efficiency of star sensors will reduce, which limiting the improvement of dynamic performance. To solve this problem, a fast star extraction algorithm based on the gray histogram is proposed. The whole star image is divided into several subgraphs of equal size, and the gray histogram and background segmentation threshold of each subgraph are calculated in parallel. Then each subgraph is marked as whether contains stars by the maximization of interclass variance (OTSU) method. Stars are extracted by the regional growth method with a certain step in these marked subgraphs. The results of several comparative experiments show that with no prior information, the extraction speed is increased by more than 60% and the extraction accuracy is improved to a certain extent. The dynamic performance of transferring to star tracking is improved from 0.8°/s to 2°/s. In conclusion, the algorithm is robust and efficient under various working conditions, thus demonstrating high engineering application value.