SOC implementation of real-time star object detection

In order to solve the problem of the image detector on low orbit satellite, and the combination of the image data flow transmission and the SOC, the Laplacian transformation and Gaussian filter hybrid algorithm to suppress the pulse noise and random noise, and then, using the method of gray weighted center of mass, and realize the real-time processing of multi-star position. This target detection algorithm can detect star targets in real time in SOC, realize the function of target centroid extraction in the image acquisition process, and the processing speed can reach more than 15 frames / s, with a good detection probability, and is suitable for the real-time requirements of star target detection in deep space background.