• Font Size:
  • A
  • A
  • A

Feature Articles

Printing Goes ‘Custom’ with the Help of Machine Vision

by Winn Hardin, Contributing Editor - AIA

Traditionally, high-speed printing applications have depended on both the trained eye of skilled machine operators to control printing equipment and controlled processes to guarantee that printed materials are of the finest quality. In recent years, however, the need for productivity, as well as an increase in the number of small-lot, customized printing applications has meant that printing press OEMs use more automation to keep the presses rolling efficiently. Machine vision has responded, helping OEMs and integrators to automate much of the printing control process, including text and pattern inspection, as well as general press control and registration functions.

Machine vision systems in printing applications come in many forms. Line scan cameras image large swaths of web-based printed materials, verifying that ads, text, and other patterns are being printed with the required accuracy.  In addition, area-scan cameras are used to verify the quality of pictures and graphics, as well as the general registration of multi-color screens used to create full-color images. The increase in small lot manufacturing has also impacted packaging and marketing materials, resulting in the need for printed materials that have unique textual and product tracing requirements. Today’s machine vision for digital printing applications means that the printed materials have to be checked not only for readable text, but for the correct number of collated pages based on the individual recipient, the correct pages in the right envelope, as well as serial barcodes for product promotions that allow manufacturers to track customers responses and habits, allowing the manufacturers to further optimize their sales. For example, credit cards with their individual names and account numbers complete with security codes illustrate a highly customized printing application.

All of these printing applications require high-speed operation without data loss or corruption with easy data interfaces to OEM and automation equipment. In response, machine vision companies have focused their product development on the sensor’s shutter and jitter control as well as image processing electronics and data interfaces to make sure their machine vision systems deliver the printed goods.

CCD’s Rule, but CMOS Coming up Fast on the Outside
Today’s printing presses are highly specialized pieces of equipment produced (generally) by a handful of knowledgeable OEMs. These presses run very fast – on the order of thousands of meters of paper per minute, ‘‘…and we need to be able to help and control the quality,’‘ explains Ender Toth, U.S. market manager and application engineer for Vision Components (Hudson, New Hampshire).

Machine vision systems using area scan cameras typically track page registration for color printing, and ribbon control, among other press functions. Careful sensor design is critical for fielding a successful machine vision-based quality control system for printing.

‘‘We need a jitter-free image,’‘ Toth explains. ‘‘At these speeds, that means you need shutter speeds as fast as 5 microseconds that can deliver high-integrity images. This typically means that printing applications use CCD cameras over CMOS. While you can window CMOS sensors and use other methods to make them very fast, thus optimizing a CMOS sensor so that a particular operational parameter is equal or better than a CCD sensor, it is still hard to build a CMOS sensor that can perform as well as a CCD sensor for high-speed, high-quality imaging applications. That being said, CMOS sensors have come a long way and continue to improve. It won’t be long until CMOS can do what CCD’s do for printing applications.’‘

To keep up with the fastest printing presses, Vision Components’ smart camera products use internal memory buffers that allow each pixel to be processed as the next image is acquired by the sensor. ‘‘The frame grabber is basically included in the smart camera along with the memory storage and processing elements,’‘ Toth adds.

For high-speed applications, some camera manufacturers may consider mutli-tap methods – essentially using multiple read out electronic circuits to read out multiple sections of the sensor simultaneously. However, in many printing applications, registration, pixel-to-pixel intensity variations and spatial accuracy are just as important as speed. Today’s printing applications may use micromarks to track page registration of multiple screens hidden in the text rather than standard registration marks in the margin of the printed page. These minute ‘micromarks’ mean that the vision system must be able to deliver a  robust, high-resolution image every time. Registering multiple sections of sensor together in a mutli-tap architecture – each with it’s own read out circuit and slightly different response - introduces additional complexity and cost into the machine vision printing verification system.

‘Extra Extra’ Considerations
Because OEMs dominate much of the printing industry, machine vision suppliers need to be prepared to support their systems for a long time – perhaps a decade or longer, while continuing to introduce new computational elements and controls to attract new projects and customers. Many machine vision equipment suppliers that target the printing industry will make sure they have data interfaces that are common to industrial equipment, using the generic TCP/IP, HTTP, or Telnet protocols, or more specialized field bus networks, such as Profibus or Modbus. Typically, valued added resellers (VAR) or integrators steeped in the special needs of the printing industry handle the final interfacing between the vision system and the OEM equipment.

Another benefit that attracts printing applications is the integration of multiple sensors, and/or display capabilities directly into the sensor housing, or when unnecessary, stripping away the camera housing to the board level to make it easier to retrofit or include machine vision into existing or new OEM equipment. Vision Components, for instance, has developed a new two sensor smart camera for the robotics industry that is finding use in the printing industry, not as a stereoscopic robotic guidance system, but to accommodate wide webs, or two-sided simultaneous inspection of the printed page.

Like many industries, coming out with new functionality often takes a back seat to delivering a robust, trusted solution, and the printing industry is no different. ‘‘A machine vision system for printing needs to be very robust because the presses generate high magnetic and electrical fields as well as a lot of motion and vibration. You need to make sure your equipment can handle the vibration and messy environment of the press, or it really doesn’t matter how smart your smart camera is,’‘ concludes Toth.



There are currently no comments for this article.

Leave a Comment:

All fields are required, but only your name and comment will be visible (email addresses are kept confidential). Comments are moderated and will not appear immediately. Please no link dropping, no keywords or domains as names; do not spam, and please do not advertise.

First Name: *
Last Name: *
Your Email: *
Your Comment:
Please check the box below and respond as instructed.

Search AIA:

Browse by Products:

Browse by Company Type: