Discriminant Study Of The Laser-induced Damage Of Optical Films

The reliable measurement of laser-induced damage threshold and improvement of laser damage defence capability of optical component are the research focus to scientists. In accordance with the international standards of ISO 11254, surface images of film are acquired before and after laser shot by a LIDT automatic testing system. Software are written to evaluate laser-induced film damage by matlab. Further more, the damage characters are extracted in high-resolution images of injured film, through which we try to get some information to conclude damage mechanism.In the first part, the methods of measuring LIDT of optical films are summarized briefly. Secondly, a LIDT texting system is built, and the system is composed of testing hardware and software systems. Every part of measuring platform and their work principle are described in the paper. Algorithm identifying damage of microscope images are studied in software of testing system, and algorithm models of film image before and after laser shot collected instantaneously by a CCD camera are built. For eliminating the noise of film image and proper partition, comparing different kind of filting method and thresholding, averaging operator, Laplacian operator and open operation reconstruction are combined to filter the noise of image, and Otsu thresholding is the proper binarization, which could extract the damage region effectively, even for the image lacking adequate contrast. The number of damage pixels, minimum out-connected rectangle arithmetic calculates the area and the diameter of damage region separately.Fundamental laser damage mechanisms of films are studied in the paper. Through classic injure mechanisms analysis, we find close connection between damage morphology characteristics and representative damage mechanisms. Supported by the damage theory and experiment results, some representative damage characters in image taken in our experiments are analysed, just like color characteristics and geometrical characteristics extracted to be standards identifying damage types. In our experiment, it is found that the shape of inclusion breakdown are not closer to circle than thermal damage and mechanical stress damage, which are shown by the result of shape parameters. Distributing difference measure of RGB average and number of connected component could make thermal damage distinguished from other damage types.Finally, we get conclusion:similarity judgement based on subtraction of images histogram-based similarity measure algorithm can identify the damage of films fast. Algorithm of surveying damage area and diameter implements quantitative detection and damage identification effectively. Color characteristic parameters and geometrical characteristic parameters of the microscope images of TiO2 single layer film are extracted to analyze qualitatively the main damage mechanisms, which may be evidence of other methods to study damage mechanisms and has significance to other types of films in studying mechanisms from damage morphologies.