Star Point Target Extraction with High Noise Background Based on Morphology

Star sensor positioning accuracy is the most important performance index of Star Sensor. However, noise is also an important factor affecting the positioning accuracy of star sensor. In order to extract the small star-point target effectively and improve the efficiency of star-image recognition and count accuracy of satellite attitude, a morphological star-point extraction method is adopted in this paper. We can get an estimate of the image background based on the characteristics of mathematical morphology operation. By using the Top-Hat transform of gray-scale morphology to cancel the background, we can get images with targets and high-frequency noise. In addition, in this paper, we also use adaptive threshold method to determine the actual image threshold. The experimental results [1] show that this method can filter background noise very well, and it is helpful to select the threshold value and determine the target brightness. The object luminance determined by this method is better than that obtained by the traditional method. It is proved that the difference between the centroid coordinates and the true values is no more than 0.26 pixel units.