基于边缘高斯拟合的过曝光星象聚心

摘要

The centroiding problem of stars that suffered by over-exposure was explored and an algorithm based on Gauss plane interpolation with data of edge region was proposed. The Over Exposed Region (OER) was detected first to determine the row (or column) data for calculation, then the interpolation of Gaussian plane with edge region data was conducted to gain candidate location result of each row (column). The final confirming of last result was then accessed by a comprehensive integration of results with an OER related weighting scheme. Simulation and results analysis proved high reliability and robust capacity of the algorithm.%对过曝光条件下的星象聚心进行了研究,提出一种基于空间曲面边缘区域信息的不完全高斯拟合算法.算法首先检测过曝光星点区域确定各行(列)参加曲线拟合的像素点,再结合高斯空间曲面模型进行边缘拟合得出初步定位结果,最后以与过曝光区域尺度有关的权重对各行(列)计算结果进行综合处理进而得出最终结果.仿真试验表明,算法具有较高的可靠性和抗干扰,能够较好的解决包括过曝光条件下的星点定位问题.