Image data set for AI-assisted reliability assessment for gravure offset printing system

Published: 31 August 2022| Version 2 | DOI: 10.17632/fpf2jv378d.2


In total, 299 images of printed lines were collected using the in-house roll-based gravure offset printing system.  Then they were labeled for overall printing quality classification and local printing defect detection tasks. For overall printing quality classification, these images were divided into two classes: 225 with satisfactory quality and 74 with defects. For local printing defect detection, all 74 images with distinctive local defects were chosen and labeled following the YOLO object detection format. During the training process, these images were split for training and validation as follows: 75/25 % for the overall printing quality classification model and 80/20 % for the local printing defect detection model respectively.



Korea Institute of Machinery and Materials


Electronics, Printing Industry