基于几何特征的圆形区域星图识别算法

圆形区域法是星图识别算法之一，具有识别过程简单但不能排除伪匹配的特点。对圆形区域法进行改进，以右手螺旋法则构造三角形模式的几何特征；通过在最亮两颗星的内外环交集内引入第四颗亮星消除三角形的相似性，减小伪匹配发生概率。对猎户座实拍图片进行识别，并对全天随机位置产生模拟星图，在不同的星体位置噪声下进行星图识别。仿真实验表明，采用改进后的圆形区域法星图识别率达到98％，识别时间平均低于0．05s。

Geometry Character Based Circle Region Star Identification Algorithm

Circle regional identification algorithm is one of the star field identification algorithms.It has a simple identification processing but can't get rid of wrong match.Importing the geometry character of the triangular mode by the right-hand rule;dispelling triangle similarity through introducing the fourth brightest star in the internal and external ring mixed region of the two brightest stars,it reduces the probability of false match.It recognizes the real map of the Orion constellation,generates star map at random sky position,then tests the identification ability at different position noise..The simulation shows that when using the improving algorithm,the identification efficiency is 98%,and the average identification time is 0.05s.