Goal-setting – how small is small?

If you’ve ever developed a defect detection system you’ll have been faced with quantifying the smallest defect that must be detected. Usually the goal is to find whatever the human inspector can see, but any engineer will tell you that’s a vague specification.

The limits of human visual acuity is a topic both interesting and relevant to machine vision, but we’ll save it for another time. Today I want to share a splendid article titled, “There’s a Hole in the Bucket,” written by Glenn Archer of Arch Consulting Enterprises and published in Advanced Imaging, (March 2009.) In an article that manages to be both technical and readable, Glenn describes a situation where he needed to set a lower limit on defect size. As he explains, it’s a complicated subject that requires much discussion with the customer. But get it wrong and there’s a strong chance you won’t get paid.

I’m not going to spoil the story by telling you how it ends; you’ll have to click through to Advanced Imaging and read it for yourself. I’d love to share a link to Glenn’s business, but the web site is still under construction.