Research On Laser Thin Film Damage Identification Method Based On Image Features

As the core component of optical system,optical thin film is mainly used to control the quality of beam propagation in optical system,so as to protect the system from damage.In the high power laser system,the ability of optical thin film to resist laser irradiation is the key factor to determine the power level and operation performance of the whole system,which is always the bottleneck to improve the laser power.Therefore,it is necessary to accurately measure the laser-induced damage of optical thin films.The microscopic method is the most intuitive method to determine whether the film is damaged or not,but the efficiency of this method is not high Therefore,in order to improve the efficiency of laser?induced damage threshold measurement and enrich the damage evaluation methods,the gray-scale feature and RGB color feature of the laser-irradiated thin-film image are analyzed and used to distinguish the damage of the thin-film.On the basis of theoretical research,the image features are studied by using on-line charge coupled element(CCD)and digital microscope.The laser damage test platform was set up,and the images of thin film samples irradiated by different laser energy were collected online by CCD.Gaussian low-pass filter is used to pre-process the image before and after laser irradiation,and the pre-processed image is differentiated.The gray value of laser irradiated region is extracted from the image which contains only laser irradiation region.The variation rule of different laser energy and gray value is studied,and the function relation between laser energy and gray value is obtained.Combining with the damage of thin film obtained by microscope observation,the gray value is calculated as the criterion for judging the damage of thin film.The microscopic image of thin film was observed by digital microscope with magnification of 100 times.The thin film microscopic image was pre-treated by Gaussian low-pass filter,.Using digital image processing methods such as binarization,the grayscale and RGB values of the image damage region are extracted,and the variation rule of the gray and RGB values of thin film microscopic images under different laser energy is studied,and the damage of thin film obtained by microscope observation is combined with the change of gray level and RGB value of thin film microscopic image under different laser energy.The grayscale and RGB critical value of the microscopic image are obtained when the film is damaged.The results of thin film image characteristic analysis show that the gray value and RGB value of thin film sample images are different when laser irradiated or not,and with the change of laser irradiation energy,the image of thin film samples collected on line by CCD is different.The gray value will increase with the increase of laser irradiation energy.In the microscopic image of thin film under microscope,the change of grayscale value and RGB component value is related to the micro-damage morphology of the film sample.TiO2 thin film and LaTiO3 thin film were tested respectively.In the sample images collected on-line by CCD,when the gray value of TiO2 film and LaTiO3 film were higher than 21.65 and 28.43 respectively,the damage of TiO2 film and LaTiO3 film would occur.In microscopic image of thin film,when the gray value of TiO2 film is less than 120,or the critical value of RGB is less than 121,122 and 120 respectively,the thin film will be damaged.For LaTiO3 thin films,the damage will occur when the gray value is greater than 170,or when the RGB critical value is greater than 165,163 and 170,respectively.The gray scale and the RGB critical value obtained by the two methods are used to judge the film damage,the results are in good agreement with the results obtained by the scattering light discriminant method.This method not only enriches and extends the method of thin film damage evaluation,but also provides a new criterion for evaluating the ability of thin film to resist high power laser.