Font Size: a A A

Print Defect Detection And Classification System Based On Image Processing

Posted on:2019-04-19Degree:MasterType:Thesis
Country:ChinaCandidate:Q Y LiFull Text:PDF
GTID:2428330566991368Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
In recent years,with the rapid development of printing industry and the continuous improvement of printing technology,the requirements of printing precision are also increasing.There is undoubtedly a higher requirement for the quality detection of printed products.However,the current printing test technology in China is in a stage of development,In view of this problem,for the sake of real-time and accuracy of defect detection in the printed matter,this paper conducts a study on the defect detection and classification technology of printed matter,which is significant for defect detection in printing industry.This paper combines the technology of image processing,pattern recognition and printing industry,and designs a set of printing defect detection and classification system,which is based on image processing technology,to realize the detection of defective print and the classification of defect types.After preprocessing the print image,the defective images will be detected and they will be displayed by the image difference method.The defect detection accuracy of the algorithm is high.In the classification of defect types,on the basis of establishing BOW model,the classification algorithm of support vector machine is used to classify the type of defect.After several experiments and comparison,the suitable classification kernel function is selected.After a number of samples training,the classifier achieves an ideal classification state.The class accuracy rate is over 90%,which meets the requirements of the system.According to the function requirements of the system,a set of software system for detection and classification is designed,which realizes the functions of system login,image acquisition,image preprocessing,defect detection,defect image display,defect classification and so on.The results of several experiments show that the software system runs steadily and the accuracy rate of defect detection and type classification is high,which realizes the predetermined target and meets the requirement of precision.Meanwhile this software system has a certain practical value for the development of print defect detection technology.
Keywords/Search Tags:image processing, printed matter, defect detection, defect classification, support vector machine
PDF Full Text Request
Related items