基于卷筒纸胶印的印品在线缺陷检测系统

AReal-time Print-defect Detection System for Web Offset Printing

本文详细介绍了一种视觉系统,该系统可以实现对卷筒纸胶印过程中产生的印品缺陷进行检测和定位。具体来说,该系统能够对高速卷筒纸印刷进行实时监控,并且可在印刷过程中出现任何影响印品质量的故障(如结构缺陷、色彩偏差、漏字、墨点、墨杠等)时向操作人员发出警报。本文还介绍了机器视觉系统的硬件部分和用于彩色图像扫描及印品缺陷检测定位的图像处理算法。从本质上讲,该系统采用高速扫描算法检测图像的边缘和边界,利用线性和非线性的滤波器对动态大小、阈值和转换信息做进一步的分析。除了在实验室内进行实验以外,还建立了系统模型,系统模型可应用在卷筒纸印刷中,在真实条件下对模型的功能进行评价。将该系统安装在凹印机上进行实验,结果表明该机器视觉系统同样可以成功地对缺陷进行监控和检测。

This paper presents a detailed description of a vision system developed to detect and locate the non-uniformities that appear on the web offset printing machine. Specifically, the system is capable of monitoring the high-speed web offset printing in a real time environment and alerting the operator of any events （e.g., structural faults, color variations, missing characters, ink splashes, streaks, etc.） that disrupt the uniformities in the web offset printing. Such events are thought to affect crucial printing properties, resulting in non-uniformities in printing and impacting its quality and printability. This paper describes the vision system in terms of its hardware modules, as well as the image processing algorithms that it utilizes to scan the color images and locate areas of defect from the printing web. Basically, the system utilizes high-speed image scanning algorithms to detect edges and boundaries using linear and non-linear filters of dynamic size, threshold and transformation for further analysis. In addition to being tested in a laboratory environment, a prototype of this system was constructed and deployed to a web printing system, where its performance was evaluated under realistic conditions. The system was installed on a Flexo gravure-print-press machine for testing, and it was found that the vision system was able to successfully monitor and detect non-uniformities.