Research On Surface Defect Detection System Of Optical Lenses Based On Machine Vision

In the electronics industry,optical instruments and even in daily life,all aspects have been inseparable from the use of optical lenses,and the industrial production lines for optical lenses are gradually increasing.However,during the processing or use of the optical lens surface,a series of defect damages are often generated,and these defects will definitely affect the performance of the element.At present,the lens companies in China have relatively limited methods for lens detection.Not only do they need to be manually operated,and the instruments are cumbersome to maintain,which does not meet the automation requirements of modern industrial production lines.Aiming at the problem of accurate recognition of optical lens surface defects,this paper proposes an optical lens inspection system based on machine vision,which can be widely applied to the defect detection of relevant optical lenses such as camera lenses and eyeglass lenses.Build a hardware platform based on machine vision technology.This dissertation designs a detection device composed of electromechanical device,image acquisition system and processing module.The design of the electromechanical device mainly includes a conveyor belt and a classification device,both of which are connected to a computer by a single-chip microcomputer.The computer controls the transmission speed or issues a stop command,and the classification device also performs the image processing result of the computer to complete the classification and screening of the lens;It is a key image acquisition system.By analyzing the light source,lighting method,industrial camera,lens and working environment,the appropriate experimental part model is selected to achieve the best surface image acquisition effect;and the processing module is a PC computer.Constitute,coordinate the overall process of inspection work,analyze defect inspection results,control the normal work of each module,etc.,and finally complete the construction of an online detection platform.The software design is based on the surface image of the optical lens,and a series of algorithms are discussed and studied.The first is the determination of image processing algorithms,from image preprocessing to image segmentation,from edge detection techniques of different operators to mathematical morphology processing,and finally,Laplacian sharpening,median filtering,and Canny-based operator are established.The edge detection is composed of algorithm programs.While paying attention to accuracy,it requires a higher speed,and the image display results prove the reliability of the theory.The second is the extraction and classification of defect features,and the comprehensive calculation of the shape feature parameters of surface defects.The decision tree classifier based on C4.5 algorithm is selected.For the various types of defects such as scratches,pockmarks,and scratches on the surface of the optical lens,the parameter set {roundness,convex area ratio,aspect ratio,circularity} was used as the classification basis to achieve higher classification.Accuracy.Finally,through the MATLAB software's GUI interface,a software system for optical lens surface defect detection was established,and the overall framework of hardware and software was expressed.This paper makes use of machine vision technology to try and explore the surface defects detection of optical lenses.It finally completes the process from the theoretical program algorithm to the actual classification work.Through the observation of a large number of experimental results,it verifies that the detection system program is feasible.