Research On Visual Inspection System For Printing Defects Of Wine Cap Printing Plate

The surface of the bottle cap is often printed with a specific pattern.Due to the damage of the printing plate and the printing process,the printed pattern may have defects such as missing characters,multiple paints and scratches,thereby affecting the overall aesthetics of the bottle packaging.At present,the detection of surface printing defects on the bottle cap mainly relies on manual screening,which is easy to cause visual fatigue to reduce the detection efficiency,and the detection rate of subtle defects is low.This thesis studies the three defects of character missing,multi-paint and scratch in the process of cap printing,and it is detected by machine vision.According to technical indicators of the manufacturer,the printing bottle cover printing plate printing defect detection system is designed.Through the design of the hardware platform,the imaging system for the printing of the bottle cap is completed.According to different types of defects,the detection algorithm is designed by using image processing software.Thereby,the overall detection system of three common printing defects on the surface of the printing plate of the wine bottle cover is realized.This topic analyzes the defect formation method,the appearance position and the characteristics expressed on the image in the printing defect of the wine bottle cover printing plate.An image registration method based on the intersection of angular edges and bilinear interpolation is proposed to detect the missing characters on the surface of the printing plate.For the multi-paint defects caused by the ink leakage or uneven printing during the printing process,a method based on the combination of regional moment feature localization and regional gray level and entropy characteristic curve is proposed,which is more paint for “small area”.Defects are detected.Aiming at the scratch defects on the surface of the wine bottle cover printing plate during storage and transportation,a sub-area detection method based on gray threshold segmentation is proposed to detect the scratch defects in the background of the bottle cap region.In order to test the stability of the system and the feasibility of the algorithm,the character missing,multi-paint defects and scratch defect algorithms on the surface of the printing plate of the wine bottle cover were tested in the self-built wine cap printing plate library,and the test results were analyzed.The analysis results show that the minimum precision of detection is 0.5mm,and the accuracy of detection algorithm reaches 93.62%.Therefore,it is verified that the printing bottle defect printing detection system designed in this thesis has good practical application value.