亚表面缺陷的检测和去除对于提高光学元件的激光损伤阈值至关重要。结合全内反射显微技术和数字图像处理技术获得光学元件亚表面缺陷信息的新方法，利用显微镜系统的有限焦深，对亚表面缺陷沿深度方向扫描，可以获得不同离焦量下的散射图像，通过数字图像处理技术，建立缺陷散射图像清晰度评价值与离焦量的关系，通过清晰度曲线得到亚表面缺陷的深度位置及深度尺寸。模拟全内反射显微平台的成像过程，讨论微调焦过程中全内反射显微成像的特点。缺陷深度位置及深度尺寸的测量精度主要由载物台精密调焦机构的精度以及显微镜的焦深决定，一般可达微米量级。利用飞秒激光加工技术制备尺寸和位置已知的微结构，使用该方法准确获得了微结构信息，验证了该方法的有效性。

Subsurface damage (SSD) detection and removal play an important role in improving the laser induced damage threshold (LIDT) of optics. A new method to detect the distribution of SSD is proposed. This method combines total internal reflection microscope and digital image processing technology. Because of the limited focal depth of the microscope system, one can get different images focusing at different depth levels by a total internal reflection microscopy (TIRM). The connection between the definition curve of TIRM image sequence and the depth position of SSD is established. Also, the depth length of SSD through the feature of definition curve is obtained. The imaging process of TIRM is simulated and the typical features of the focusing process is found out. Size information of the microstructures is got exactly according to the fabricated microstructures by femtosecond laser micromachining. Focal depth of the microscope determines the test depth accuracy of our method and actually the test accuracy can achieve 1 μm.