A Theoritical And Experimental Study On Surface Defect Detection Of Soft Contact Lens

Contact lenses are widely used in the fields of vision correction,beauty and makeup,eye disease treatment,smart sensors and other medical devices.It is particularly important to detect the surface defects of contact lenses,which will cause irreversible damage to users.However,due to the complexity of production process,the surface defects are inevitable.Actually,the surface defect detection mainly relies on manual work now due to some detection difficulties such as special material,complex characteristics of defectes and the interference of foreign impurity.Therefore,automatic,intelligent and efficient detection technology,especially the research on core detection algorithm,has become the key technology and urgent demand of equipment research and development in this field.In this paper,a soft contact lens surface defect detection method is proposed,which is based on the production demand of enterprises and the author’s quality testing experience in the production line.And this method has been successfully applied in the lens surface defect detection of contact lens production line.Firstly,this paper introduces the soft contact lens research progress of surface defect inspection technology.Then,based on the structure of contact lenses and the author’s own production experience,the causes and morphological characteristics of surface defects are analyzed,and the image processing principle of surface defects detection is systematically discussed.Further,in this paper,the soft contact lens surface defect detection system is built with theoretical analysis and practical experience including the design of the imaging methods,the design of automated image acquisition device,the design of image acquisition software as well as the basic image preprocessing operations,including the area of interest extraction,color correction,gray scale normalization,etc.,In the part of color correction,the method of color correction by creating "gray world" is proposed.Finally completed pre-processing image design the specific defect detection algorithm,mainly including was proposed based on the statistical characteristics of gray level of detection algorithm,edge extraction algorithm based on the preset sampling box,decorative pattern extraction algorithm based on color channel separation,decorative pattern elimination algorithm based on frequency domain texture filtering,and lens internal defects detection algorithm based on blob analysis.In addition,some experiments are carried out to verify the proposed detection algorithms.The defect detection rate experiment of manual reinspection was carried out,and 1784 defect samples were detected,the detection rate was 97.59%,realizing the rapid and accurate detection of contact lens surface defects.The method of soft contact lens surface defect detection proposed in this paper provides reliable technical support for automatic,intelligent and efficient detection of soft contact lenses.