The On-line Detection System Of Printed Matter Based On Machine Vision

With the rapid development of economy and the advancement of science and technology,more and more prints have appeared in people's daily lives.However,due to immature printing technology and other internal and external factors,printed defects caused by defects have occurred.In particular,with the increase in printing speed,traditional inspection methods have obviously failed to meet the requirements of the modern printing market.The development of machine vision technology has brought a gospel to the realization of the defect auto-detection technology.The regret that could not be achieved in the automatic detection of print defects has also been resolved due to the progressive development of machine vision technology.The detection system(online detection defect)designed in this paper is a typical representation of the machine vision system framework.The detection system is composed of the acquisition image card,the corresponding light source,the computer(PC),the camera,and the execution control module.Then,this article introduced the basic framework and ideas of how to detect defects,and also elaborated some corresponding specific knowledge of image processing.It mainly expounded the concepts and common methods of image preprocessing.Finally,it introduced the concept of binarization and Common method.Several flaw detection algorithms have been deeply studied and compared.The goal is to make the code simple and the system good real-time.Analyze the advantages and disadvantages of several defect detection algorithms,and finally determine the algorithm used to detect defects in this system.The technology used is the image difference method.At the same time,comprehensively analyze the defect detection process.This article uses VS2010 to design a software system based on the MFC interface,and introduces the main functional modules of the software according to the development sequence.At the same time,an intuitive display of the common defect type detection results.Finally,the real-time performance and detection accuracy of the system were evaluated.Through experiments,the online defect detection system based on machine vision for this problem is able to detect the common defects such as scratches,ink bars,and stains,which meets the needs of defect detection.