Research On Ink Control In Ink Area Based On Machine Vision

The amount of ink directly affects the quality of printed matter.Therefore,the rapid and accurate detection of ink quantity information is very important to improve the quality of printed matter and reduce the cost of printing.With the progress of science and technology and its integration and penetration,about the detection and control of ink,the traditional subjective detection and control method based on human eye observation has completed its mission,and the method and technology of ink detection and control in ink area based on machine vision has become an inevitable trend.On the basis of previous research,this paper has carried out systematic analysis,core technology research and experimental analysis on the detection and recognition of ink area ink based on machine vision,(1)Aiming at ink detection in ink area based on machine vision,systematic method research is carried out,and the basic ideas and methods of ink detection are putting forward.The design determines detection flow of the system,and the color space conversion m,ethod and the conversion precision evaluation analysis is carried out for the core module of the system development.Based on the color target scanning image,the chromaticity detection method is adopted,the polynomial model transformation method is selected as the color space conversion model,and the CIELAB colors difference formula is used as the basis of color difference evalu,ation to determine the parameters of different color space conversion models,and the color detection method is used to determine the color space conversion model parameters.And according to the ink area overall chromaticity calculation and analysis to realize ink area ink monitoring.(2)Using Adobe Illustrator CSS designed platform,the RGB experimental color targets file,CMY experimental color target file and test sample sheet was designed and made.Through actual printing,the RGB printed color target,CMY printed color target and experimental test sample sheet were obtained.Selecting the printed color target sample sheet,the printed color target was scanned with EPSON Perfection V850 Pro scanner,and the electronic image was obtained.The RGB of color targets color block was read by MATLAB program,and the L a b*value of the color targets color block was measured manually by X-rite eXact spectrophotometer.The dot increment value of the trichromatic ladder ruler is measured,and the set CMY value is corrected.(3)Taking MATLAB2016b as the platform and using the interface design with good function of GUI,a set of ink detection system based on machine vision is developed.The system has the basic functions of reading RGB data,calculating the model parameters and analyzing the accuracy of the model.The predetermined function of ink area ink quantity detection is realized,and the result of ink area ink detection of standard sample sheet can also provide reference for printing press ink area ink preset.(4)Based on the printing experiment data,the experimental examination and analysis was conducted.Firstly,the polynomial model coefficients of different order are solved,and the conversion precision is calculated by color difference analysis.Secondly,through the precision comparison,the model optimization and parameter setting are completed,and the RGB-CIEL*a*b*colors space conversion model and the CIEL*a*b*-CMY conversion model of the scanner is determined.Finally,directing at the test data of color target,the ink quantity test in ink area is carried out.The test results show that the amount of ink in some sample sheets is normal,the amount of ink of some sample sheets needs to be adjusted,and the ink area and the ink adjustment reference value for adjusting the ink are given.The feas.ibility of ink area ink detection system based on machine vision is verified.