【摘要】 讨论了一种新的无须任何先验知识的星图识别算法。该算法将四边形星图模式分解为两个具有公共边的三角形模式 ,使用三角形模式的特征 ,在保证算法计算量和星载星表的存储容量的前提下 ,使算法的识别成功率得以星著提高。 Monte Carlo实验表明 ,在星等误差0 .5星等 ( 3σ) ,位置误差 1 0角秒 ( 1 σ)时 ,该算法的识别成功率接近 1 0 0 %。更多还原

【Abstract】 A new star pattern recognition algorithm without any a priory attitude knowledge is described.The algorithm divides a star pattern,which consists of four stars into two triangles with a common side.Using the features of triangular pattern,the success rate of the star pattern recognition scheme increases obviously,while the recognition time and the storage capacity of recognition processing keep almost unchanged.Monte Carlo experiment demonstrates that,while the magnitude uncertainty is 0 5(3σ) and the position uncertainty is 10 arcsec (1σ),the success rate very close to 100%.更多还原