Research On Image Detection Technology Of Printing Quality

In the context of the transformation and upgrading of the manufacturing sector,the printing industry is facing the problem of traditional technological transformation.In order to effectively improve the degree of automation of printing quality detection,on the basis of machine vision and digital image processing theoretical knowledge,the automatic image detection technology for printing quality is researched.Through the use of modular design,the printing detection experiment system is realized,which provides a feasible scheme to solve the problems of low efficiency and poor stability in the actual printing process.From the two aspects of system composition and algorithm design,the application of printing quality inspection system based on machine vision in industrial practice is introduced.The main research contents of this paper are as follows:(1)According to the actual needs of production,reasonably determine the selection scheme of the hardware equipment of the system,a printed image acquisition device is composed of conveyor belt,photoelectric switch,rotary encoder,LED source,CCD industrial camera and the PLC controller.In the collection of printed images,the system uses the Pylon Viewer program to drive the camera to automatically complete the accurate shooting of the print.In the light of the influence of the external illumination on the image acquisition process,the device is used for sealing the camera and the light source.(2)In order to ensure the smooth implementation of the system of printing image automatic detection process,this paper puts forward a series of printing image processing and recognition algorithms.The key algorithms of image grayscale,image enhancement,image segmentation and so on are introduced to ensure the accuracy of subsequent test results.Two kinds of printing image registration algorithms based on ROI template and Hough and Fourier transform are proposed,which lays a good foundation for further defect identification.The algorithm of defect target extraction and classification is designed and the multi-class support vector machine is used to recognize and classify the defects.Finally,based on the C++ programming language,using Visual Studio development tools,by the use of programming and software project development knowledge,the printing quality inspection system visualization platform is realized.