Smile, The Cameras are Coming

E-mail Ray Peacock

Video cameras seem to be everywhere. The city of Philadelphia has installed video systems at intersections with high incidence of traffic crashes—systems that trigger $100 tickets for traffic-regulations violators. And hardly any law or police show on TV today is complete without a suspect being identified and tracked with the help of security cameras. For automation, the camera drive is being spurred by innovations in hardware, software, and communications. The reverse is happening, too: More applications create more impetus to innovate.

The Technologies
The camera companies are having a great run of business and innovations are coming rapidly. One of these is the CMOS chip which, when paired with an anti-blooming filter, provides images free of the saturation seen whenever a camera views a very bright light source in an otherwise-moderately illuminated scene. While the CID camera has been around for many years, its use has been limited, mostly because customers had to work so hard justify its higher price. Then, too, there was only one vendor; now there are many more choices.

Machine vision is a relatively mature technology adopted by many companies in support of both discrete part and continuous processes. Machine vision systems are based on either ordinary video cameras in ruggedized housings or purpose-built cameras, rugged by design, with their own cooling and lens-cleaning. With smart cameras and smarter software, they aid in counting, sorting, classifying, identifying, validating and measuring moving products, even those traveling at very high speeds.

Cameras equipped with wireless modems can be placed almost anywhere and send video signals to a processing unit. Some intelligent cameras, like Cognex's new Checker, can perform some or all processing internally and just send back accept/reject signals.

Cameras come in many different flavors: Ethernet, Firewire, IP-enabled, miniature, high-temperature. They're available with low resolution, super-high resolution, and variable frame and integration rates. Wavelength varieties include short to ultraviolet; visible; near IR; and short-, mid- and long-wave IR.

In fact, the choices are nearly mind boggling. No longer do you need to ask whether a technology is available; if you need it, it exists—even if you don't know what you need.

Help and Why You Need It
And if that's the case, or if you don't know how to optimize your system, there are consultants, custom engineering firms, training firms, and even online courses to help. The Machine Vision Tech Group of the Society of Manufacturing Engineers provides a great forum to learn and share with others to avoid re-inventing the wheel.

Thank goodness for all these resources, because implementation isn't necessarily straightforward. For instance, lighting plays a key role in most applications' success or failure; and it's not a piece of cake (newbies walk carefully!). There are many interlocking factors, but the industry at large has learned from the people who previously have been in situations similar to yours.

Automation Applications
A feature role of video technology in automation is remote monitoring of a process area and process line (See Data-Linc and MOBOTIX AG). If something goes awry, a smart camera system can automatically activate an alarm or phone or page a maintenance alert to the proper person from thousands of miles away.

The cameras and their processing software play significant roles in both process automation and quality assurance. It's relatively inexpensive to use such systems to perform 100% part inspection when hundreds and even millions of pieces fly past the camera. National Instruments' Web site demonstrates applications of both visual and IR imaging.

FLIR Systems recently began selling a battery-powered system that uses a specially filtered mid-wave IR camera for preventive maintenance in petroleum and chemical processing plants and natural gas distribution centers. Called ThermaCAM GasFindIR, it can detect methane and other hydrocarbon-gas leaks at a distance.

A technical presentation at the Machine Vision Web site describes the use of a near-IR imager to detect the presence of methanol flames, while another shows that similar technology can detect the presence of lubricant residues on metal surfaces.

With the increased impetus to innovate, we can expect more. Stay with me and I'll update you when the time is right.

A Brief Return to RFID
A couple of months ago I wrote about RFID technology. Now, the June 2006 issue of Consumer Reports carries an article (though the full text is available only to subscribers) on how the technology came to be and why it is being adopted so rapidly in ordinary commerce.