Study On Defect Inspection Technology Of Printing Platemaking

Before the printing roll is transported to the printing plant from the plate making plant,the last process is to be passed:the page inspection.First,the printed sample is printed to reflect the quality of the printing plate by detecting the quality of the printed sample.At present,most of the testing methods for printing samples are based on the human eyes,which have the disadvantages of low efficiency and high cost.In this paper,taking the defect detection project of Yuncheng Plate Making Group as background,we set up a printing defect detection system based on machine vision to achieve automatic detection of printed matter.The specific work is as follows:1.Building the hardware platform:Based on the traditional machine vision platform,we improve the camera,lens,light source,sensor,controller and other components.Then the platform is built according to the principle.A clear image of the printed sample can be obtained by using the camera;2.Image algorithm design and improvement:through the matching of template image and the printed sample image to be matched,defect detection and defect classification,we obtain the defect information.The SURF feature point descriptor is improved to improve the performance of the algorithm.At the same time,the GA-BP neural network framework is constructed to classify the defects accurately;3.Implementation of the software platform:Using VS2013 software platform,with the help of OpenCV 3.0 and camera's own software development kit,we completed the development of print defect detection system software and realized friendly interaction with operators.The printed defect inspection system can complete the accurate detection and location of printed defects,which solves the problems faced by enterprises and has strong practicability and high utilization value.