A detection and track algorithm for moving dim targets in deep space

Analyzing the space image character, a new detection and track algorithm was brought out for moving dim targets in deep space. First, gray-scale histogram threshold segmentation was used to suppress background. Second, the morphological top-hat transformation was used to intensify the target in spatial domain, and the image difference algorithm was used to detect the target in temporal domain. Then, the both results were fused to segment and pick up the target. Finally, according to the characteristic of the candidate target, target association and tracking were used to accomplish target detection. Practical application approved that the algorithm perfectly meets the requirements of the dim moving target detection and recognition in deep space.