Latest question:
How do you think the new GigE standards will influence the machine vision industry?
Respond or ask your question now!



Two decades ago, when he founded Mercron, Ken Zeiler learned that his machine vision customers were searching for answers to their illumination needs. Most of the time, they didn't know the questions. Literally and figuratively, it was his job to shine a light on their applications. Today, the job is a little easier because of experience and the existence of knowledge-packed data bases.
"One of the biggest problems is the fact that commercial lamp literature is written for commercial and industrial uses," Zeiler says. "Take those same parameters to machine vision and they don't fit. Industrial light is based on the human eye. With a camera, it won't be as good. These are the kinds of details engineers have to make sure of. The hardest thing for an engineer is to figure out the light needs."
Too many machine vision systems developers don't spend enough time planning their lighting needs, says William Biederman of Vision Light Technologies, the U.S. representative for France-based Phlox. "People who are informed figure out lighting and presentation of the object before picking a machine," he says. "You should spend 80 percent of the time on lighting before you can use your application. I've seen people pick out their camera, lens and software, but hadn't figured out how to light the thing."
Andy Falconer, director of V Cubed in the U.K., agrees. "If you get the illumination right it simplifies it and makes sure you get the right information. Illumination highlights what you want and makes it work for you."
Mistakes
"Illumination is a crucial aspect of any application. New users need to spend time selecting illumination. It can be the cheapest and simplest, but not always. How many people make penny decisions on dollar machines?" asks Patrick Varley of Volpi.