精密表面疵病检测美标数字化评价实现算法

采用美国军用标准MIL-PRF-13830B（简称美标）对光学元件表面疵病进行数字化评价，其难点主要在于美标中表面疵病密集度的判定。针对该难点，提出了一种表面疵病密集度判定算法。该算法采用权重域叠加的方法，通过为表面疵病赋予相应的权重域来确定疵病密集圆域，并将权重域间的叠加转换成了矩阵间位置关系的判断和计算，有效的利用了矩阵运算的便利性。实验中通过对已知表面疵病信息的石英标准板进行判定，其得到的圆域即为疪病密集圆域，与预期结果一致，验证了该算法的正确性。目前该算法已经运用于光学车间精密光学元件表面疵病的美标数字化评价。

Digital evaluation algorithm of American standard in defectsinspection of precise surface

The difficulties of digitally evaluating surface defects of optics using American military standards MIL-PRF-13830B lie in judgment of surface defects concentration. For this problem, we raised a kind of judgment algorithm of surface defects concentration. The algorithm uses the method of overlap of weight region. It seeks circular area where defects are dense by means of attaching weight region for surface defects, and turns overlap of weight region into position judgment and calculation between matrixes, which makes full use of convenience of matrix operations. In experiment, using this algorithm judged quartz standard board where information of defects was known, and the circular area where defects were dense was accordant with expectation, which validated the correctness of the algorithm. At present, the algorithm have been used for precise surface defects digital evaluation using American military standards in optical workshop.