High-speed remote visual monitoring of roll-to-roll (R2R) printed packages

Abstract

Abstract— Roll-to-roll (R2R) processing is a widely adopted method in the industry, especially in the label and packaging sector. The processes involving R2R equipment are typically conducted at very high speeds, which makes it preferable. However, high-speed processing poses challenges in monitoring the products throughout the process.

In this study, a high-speed monitoring system using computer vision, capable of working remotely via an IoT module, is developed specifically for R2R manufacturing in the packaging industry. Here, the developed monitoring system is used to detect and recognize the alphanumeric series printed on the packages in the R2R production line. The series are the promotional winning codes, the system needs to capture them, recognize the characters, and compare them with a library of available codes for controlling the process. The system should verify that all codes are registered, there are no repeated codes, and the broken codes should be marked to remove and print again. For this purpose, a method of synchronizing the camera and rapid movement of the web (printed packages) is proposed. The camera captures the code on each product just in time when it passes through. The embedded OCR (optical character recognition) program inside the camera controller detects the characters and sends them for remote comparison and further processes. In this way, the system can work effectively with the linear velocity of the web up to 100 meters per minute.

The successful implementation of the high-speed remote monitoring system paves the way for various applications, as well as opens a chance for further improvements and development of in-house technologies.