Research On Vision-based Surface Defect Detection Technology Of Optical Components

In recent years,the traditional methods for detecting surface defects of optical components are manual methods.In order to improve the level of product surface quality control,and to meet the requirements of production for automation and informatization,it is of great significance to explore and study inspection methods based on machine vision.This thesis takes the most common circular optical lens in the optical system as the research object.It mainly completes the detection of broken edges of optical lenses and the detection and classification of surface scratches and pits.First,the overall design of the defect detection system is made,which mainly includes two parts: hardware and software.According to the actual needs of image processing,the hardware selection is completed,the stability of the image acquisition system is ensured with a reasonable layout,and the software processing flow of the system is briefly introduced.Secondly,it introduces the broken edge detection process of optical lens.It mainly includes circle detection and edge detection,and combined with the characteristics of the lens itself,some improvements have been made to the circle detection and edge detection methods,which can remove the interference of the internal details of the lens,and only retain the outer edge contour of the lens.The last is the detection and classification of surface scratches and pitting.It mainly includes image enhancement,filtering and denoising,ROI area selection,defect segmentation and defect classification.By combining adaptive threshold operation with morphological operation,it overcomes the shortcomings of adaptive threshold segmentation that detect the boundary information between gray levels when there are multiple gray levels in the image.Using the relevant algorithms and processing procedures researched and discussed in this article,image processing and detection classification of different optical lenses with defects have been achieved,and good results have been achieved,and the expected goals have been basically achieved,which is the realization of the subsequent engineering of detection equipment Lay a good foundation.