Research Of Screen Printing Template Measurement And Defect Detection Based On Machine Vision

In recent years,with the continuous development of China’s manufacturing industry and the gradual improvement of economic technology,more and more industries and enterprises have put forward higher demands for products.Screen printing technology plays an important role in many fields of modern industrial production,and the size defects and surface defects of screen printing template will seriously affect the quality of screen printing products,which directly leads to low qualified rate of products.At present,the detection of screen printing template basically depends on manual detection,which has low detection efficiency and strong human subjective factors.Therefore,it is of great research significance and application value to efficiently and accurately measure the size of the screen printing template and detect the defects on the surface of the screen printing template.In this paper,the visual inspection of the screen printing template is studied,and a system for measuring and defecting detection of screen printing template based on machine vision is designed.Firstly,from the perspective of dimensional measurement and defect detection,the research status and development trend of machine vision technology at home and abroad are introduced.Then the problems in the development of domestic screen printing template detection technology are analyzed,and the background and significance of this research are clarified in this paper.Secondly,the characteristics of the screen printing template are analyzed in this paper,the hardware part of the screen printing template inspection system(industrial camera,LED light source,lens,gigabit network card,computer and camera light source test bench,etc.)are studied.A hardware system that can meets the test requirements and standards of the screen printing model was established in the laboratory to collect actual screen printing model images.Then,to solve the problems of various types of screen printing templates and the objects to be tested,a method for measuring the size of screen printing templates based on machine vision are proposed in this paper.Using a positioning rectangular frame to establish the measurement coordinate system is the traditional visual inspection.A measurement strategy from thick to thin are adopted to reduce the measurement error introduced by the traditional visual inspection method.In the measurement stage,the layered matching algorithm based on the combination of the image pyramid algorithm and the normalized cross-correlation function is used to achieve the rough positioning of multiple targets to be measured,and then the threshold parameters obtained from the template information statistics are used for fine edge positioning to establish local Coordinate system to complete the final measurement.The experimental results show that the absolute error of the measurement results can be kept within 0.1mm.Finally,this paper studies the key algorithms of surface defect detection and classification recognition.For the two typical defects of the edge fracture and the incomplete roundness on the screen printing plate,the corresponding detection algorithm is designed.For the common scores,scratches,stains and inclusion defects on the screen printing template,after feature extraction,the features with strong ability to identify defects are selected according to the Fisher ratio.Compared with the BP neural network,SVM,and K-nearest neighbor method for the classification of typical defects,the SVM algorithm has the highest classification accuracy for typical defects on the surface of screen printing template.The experiment shows that the average classification accuracy is 93.4%.