Research On Digital Printing Defect Detection System Based On Machine Vision

In the digital printing industry,the detection of printing defects is an important part of ensuring product quality.Usually,manufacturers use manual visual inspection to detect printing defects.However,manual detection tends to increase labor costs,and the naked eye is fatigued,and the detection efficiency cannot be maintained.Therefore,in order to replace manual inspection,this paper develops a digital printing defect detection system based on machine vision for printing defects in the process of digital printing production.After summarizing and analyzing the characteristics of digital printing defects,the algorithm of the digital printing defect detection system is designed.The algorithm is roughly divided into three parts.The first part is image preprocessing.The image is processed through grayscale,gamma transformation and Gaussian filtering to balance the image and reduce the influence of noise on the image;the second part is feature point matching and printing based on second-order neighborhood difference Defect detection algorithm,through corner detection and feature point matching to obtain the information of the matching point pair,and then perform the second-order neighborhood difference to extract the defect information,and determine whether there is a defect according to the relationship between the connected set area and the set threshold;the third part is printing Defect grade evaluation is achieved by constructing a feature vector of printing defects and a multi-classification model based on SVM(support vector machine).The machine vision platform of the digital printing defect detection system is mainly composed of CCD area scan camera,lens,LED light source,GPU processor and other hardware.Among them,the model selection and calculation of key parameters of key hardware such as the camera,lens and light source are carried out to ensure that it can meet the detection requirements.And completed the design and development of the software of the digital printing defect detection system,mainly including the design of the software function and the development of the human-computer interaction interface based on Qt5.The system's machine vision platform,system software,and detection algorithm were tested separately.The test results showed that the machine vision platform was running stably,and the collected images could meet the needs of detection;the system software was running smoothly,and the interface function reached the expected goal;this system The average accuracy rate of various printing defects detection is above 95%,and the detection speed can reach 24m/min,which can meet the actual digital printing production detection needs.41 pictures,23 tables,66 references.