Robotics Online - RIA - Robotic Industries Association
Case Studies

Fast Forward for Image-Based ID Readers

Cognex Corporation

VSoC-powered readers increase read rates, reduce downtime and speed setup

In the not-so-distant past, manufacturers using laser scanners and conventional area-scan image based readers had to put up with inherent technology limitations. Laser scanners, for example, have a hard time reading barcodes through plastic shrink-wrap, barcodes printed on flexible material such as plastic bags, and damaged barcodes. When reading 2-D barcodes, manufacturers had to turn to conventional area-scan image-based readers, which come with depth of field limitations that make it difficult to read codes without controlling the distance between the barcode and the reader. 

Thanks to next generation area scan image based readers powered by Vision System on Chip (VSoC) technology, those days are over. With its high speed, large depth of field, extremely high read rates, the ability to handle damaged codes, omnidirectional reading, and multi-code reading ability, as well 2-D symbol reading capability, this next generation of image-based readers is poised to make life easier for manufacturers. 

Manufacturers rely on laser scanners and conventional image-based readers
In the past, manufacturers have had two main categories to choose from when solving ID challenges: laser scanners and conventional image-based readers.

Laser scanners use a moving pinpoint of light to illuminate the barcode. A single photocell receives the reflected light and converts the barcode into an electrical signal as it moves across a barcode.  The scanner then measures the relative widths of the bars and spaces, translates the different patterns back into regular characters, and sends them on to a computer or portable terminal. Figure 1 shows the basics of laser scanners.

Figure 1 - Laser ID Scanners

Most laser scanners sweep the laser beam horizontally using an electronically controlled mirror. Laser scanners tend to be quick and precise and can often read denser barcodes than other technologies. A primary advantage of a laser scanner is depth of focus; since a laser beam diverges very little with distance, scanners of this type generally have a working range from roughly 1 in (2.54 cm) to 12 in (30.48 cm). By increasing laser power and narrowing the angle of beam sweep special long-range scanners can read at ranges as long as to 30 to 40 feet (9 to 12 m). Since the laser beam is swept horizontally at a fixed angle, the length of the scan line on the target increases as the distance increases; if a barcode is too wide for the laser beam, the operator can just back up a little bit and read the barcode. 

2-dimensional laser scanners sweep the beam horizontally and vertically at the same time, creating a raster pattern. This type of scanner is used to read 2-dimensional barcodes like PDF-417. Interlocking pattern scanners use mirrors to create a grid of scan lines. Some interlocking pattern scanners use a rotating holographic disk instead of mirrors to aim the laser beam. This type of scanner is often mounted above fast-moving conveyor belts to read barcodes on packages. With a large faceplate area, the wrap-around effect makes it possible to see barcodes whether they are on the leading side, top, or trailing side of the package. Depending on the size of the scanner and the boxes it may even be possible to read barcodes on the sides of the boxes.

Although fast, low cost and with a large depth of field, laser scanners have moving parts that can wear out, have limited ability to read damaged barcodes or codes covered in plastic shrink-wrap, and cannot read 2-D codes like Data Matrix and QR. They are best for manufacturing applications reading only 1-D codes, where no trigger is available, or if part distance varies. 

The other option for manufacturers with ID applications is to look toward image-based technology or an area-scan reader. The major component of the system is camera with a CCD or CMOS sensor that detects light. Such sensors consist of a two-dimensional pixel area that is available in various resolutions. The optical lens system transfers the image onto the camera sensor. 

Intelligent embedded cameras such as vision or ID readers also contain all components needed for image processing. They also feature integrated lighting and a communications interface to a controlling PC so that you can, for example, send off barcode strings to a controlling PC or PLC. The camera field of view is the area that the camera actually can see.

Figure 2 shows the conventional image-based ID reader. Note the horizontal width and a vertical height, which defines the scene the camera can see. The camera field of view gets wider as you move away from the lens. Of course this all needs to be in a certain focal range, but with an imager-based system, the farther you can get from the barcode, the larger the area you can scan. 

Figure 2 - Conventional Area-Scan Reading Technology

This technology has high read rates and can handle damaged codes. It features omnidirectional reading and has multicode reading ability. It is, however, slow compared to lasers (about 45 decodes per second) and has a limited depth of field.  It is most effectively used when a trigger is available, where part presentation is controlled, for reading 2-D codes or mixed symbologies, and for direct part mark (DPM) applications.

Next generation area-scan reading technology – faster, more accurate, easier set up
In the conventional image-based reader, the imager, analog-to-digital (A/D) converter, and digital signal processor are discrete components interconnected via narrow communication buses across printed circuit boards. In the next generation area-scan reading technology, all of these components are integrated. The result is a step change in speed, allowing the acquisition of images at up to 1,000 frames per second, processed in real time so the reader can adapt to wide variations in package size. Figure 3 is an illustration of the new technology.

First on the market with this new technology is Cognex’s DataMan 500, which features high speed, ultra-fast auto exposure, and trigger-free reading. It decodes 90 1-D barcodes per second, twice the speed of a conventional image-based system. The system reduces overall setup costs, deployment time, and system complexity compared to conventional image-based options. It also has an increased depth of field, with large pixel size and a high speed variable focus lens. Finally, it offers reading robustness, with high read rates and a minimal need for external lighting and precision alignment. 

At the core of the powerful new technology is VSoC (vision system on a chip) technology, which provides a very tight feedback loop between image acquisition and the front end processing performed by the finder.  This means there is more time for running an intensive decoding algorithm, which ultimately results in higher speeds. 

Figure 4 compares the total cycle time between a conventional reader and the VSoC-optimized DataMan 500, showing that the VSoC optimized system features double the speed.

The system design features a reduced lens aperture size. The VSoC design incorporates large pixels and intense LED lighting allowing for smaller aperture settings. With a standard C-mount lens, the DataMan 500 can achieve 10-inch or greater depths of field. VSoC optimizes image brightness at high speeds. It works well on fast-moving parts, ensuring the best image, regardless of working distance.

Communications and the ability to connect with the Ethernet and industrial protocols is another key differentiator of these next generation systems. For example, the DataMan 500 offers Power over Ethernet (PoE) technology, which simplifies wiring in the system. PoE reduces cost by eliminating the need for a local DC power supply and individual 120V power drops at the scanning location. For seamless connection to major PLC brands, the unit provides EIP Ethernet with AOP and PROFINET.

Which manufacturing applications benefit from improved performance of next generation area-scan image-based readers?
The new technology is ideal for many manufacturing applications. One example is the food and beverage industry, where the technology helps with UPC code reading, offering high speed 1-D reading with no need for a part trigger, the ability to read poorly presented codes, and 2-D code reading for allergen applications. 

Another good example is Code 128 reading in the pharmaceutical industry, where multiple barcodes need to be read quickly, and labels with text and logos could create confusion. Here, unit dose drugs are packed in cases and the operator needs to scan all the barcodes to register serialized information and be sure the correct drugs are in the cases.  High read speed means quick scanning of many codes. In addition, reread suppression ensures no double reads. The fact that no trigger is required means barcodes can be read even when an operator is presenting the codes. 

A final example is the automotive industry, where Data Matrix codes, DPMs and varied part geometries often pose a challenge. Ultra high speed auto exposure and high frame rate means more views of these complex codes. The system’s ability to read poorly presented codes is also a plus in this application.

It all adds up to simplicity, easier set up and cost effective operations
Whereas other image-based scanners require a great deal of set up to get the distance, focus, exposure and lighting correct, the VSoC technology within DataMan500 makes the system simpler to use because position is not critical due to the large natural depth of field. The system eliminates the need for sensors, making set up simpler. It reduces system downtime and changeover time, and will result in fewer parts rejected and less hand sorting. And that’s an outcome that will be prized by many in the manufacturing sector.

Automated Imaging Association Motion Control Association