A Fast Centroiding Algorithm for Digital Sun Sensors for Nanosatellite Applications

This letter proposes a fast centroiding algorithm for Euler angles calculation based on digital Sun sensors aiming at nanosatellite systems application. The two-step algorithm uses light-spot profiling to generate columns and rows vectors by processing the image as a pixel matrix. The
x
–
y
centroid coordinates are defined by searching for the vectors indexes with the largest values. The algorithm was compared to the conventional basic centroiding thresholding method (BCTM), the multiple-threshold averaging approach, and the double balance centroiding algorithm regarding the processing time and accuracy. The proposed algorithm reduces the processing time by around 50% with similar accuracy to the BCTM approach while overcoming the other algorithm results. Later, the algorithms were optimized by introducing an image segmentation procedure to improve the initial version, in which our proposal has proved a faster performance.