6 Pages

A star crude extraction algorithm based on shearlets and top-hat transform

WithJ. Li, P. Guo & C. Wang

Starlight attitude determination generally includes four parts of pre-processing star, star target extraction, star pattern recognition and attitude solution. The star target extraction is an important part among them, and its extraction precision and accuracy will affect the performance of star sensor. Star target extraction consists mainly of two steps: star crude extraction and sub-pixel extraction. Because the stars represent the spot generally in the dark background on the CCD plane and the diameter of spot is 3-5 pixels, star point targets of star image are easily affected by noises. So it is an important factor to influence the accuracy of centroid localization that whether the star crude extract can maximize the reduction of the actual distribution of pixels and gray value of star point or not [1].