Application tip for IR lighting

Spotted an interesting little application story on the Vision Systems Designs website a few days back, “Smart cameras solve a pressing problem,” (Andy Wilson, November 1^st, 2011.)

This relates to the use of two Dalsa BOA smart cameras for checking that a die is clear before it closes. As the story relates, closing the die with slivers of stray metal in the way can cost a lot of money.

This seems like a pretty run-of-the-mill machine vision application but there are a couple of interesting aspects. Number one is that it’s the first time I’ve read about the BOA cameras actually being put to work. Number two is the use of IR lighting.

I don’t wish to steal Andy’s thunder, (or his ad revenue,) by giving you all the details here, but trust me, the application needed infrared illumination.

Now here are two tips for when you use IR. First, remember that silicon has low quantum efficiency at wavelengths much over 750nm (and goes to zero about 1,000nm,) so you’re going to need a lot of light to make the application work.

Second, most cameras have an IR cutout filter over the sensor. You’ll need to remove this in order to detect the IR light. This might seem obvious but I have found vision engineers complaining that an IR light isn’t working when all they had to do was remove the filter!

Just a couple of points to bear in mind.

What could infrared imaging do for you?

If you didn’t “get” the witty visual gag I shared with you in my last post, “Bringing out the physics geek,” let me explain. The parking spaces were numbered in wavelengths of light. Thus the green car was in space 532, the yellow in 561 and so on. Space 1054 appeared empty because we can’t see light of that wavelength: it’s infrared.

But what would we see if we could? Well Sensors Unlimited, who specialize in short wave infrared cameras, have some great examples on their web site. A few minutes browsing through their gallery might give you some ideas for that particularly challenging inspection problem you’ve been chewing on for months.

Now, before you pick up the phone to find out just how expensive their fantastic cameras are, let me offer another suggestion. Most regular machine vision cameras have some sensitivity to IR light, (if yours has an IR cutoff filter over the sensor try removing it before playing with IR illumination,) so get your hands on an IR security light and find out what you can do. This won’t get you in to the real SWIR part of the spectrum, but it might be far enough along to help with fruit, glass or plastic inspection.

Don’t forget to let me know of any successes you have with near IR imaging!

See what you can do with IR!

No sooner had I finished penning my last post on the merits of IR illumination than I stumbled across another piece explaining how long wavelength light can help with machine vision tasks.

“IR microscopy helps boost 3-D IC yields,” (Test & Measurement World, October 2010,) addresses the way in which IR can help with inspection of silicon products, such as PV cells. The key here is that silicon transmits IR light, which means an IR backlight can give a kind of x-ray effect.

Check out the article; it might give you some ideas for your next machine vision project.

The influence of wavelength

Near IR illumination is interesting, not just because it highlights things that we can’t see with our eyes, but because of the ability of longer wavelengths of light to penetrate materials that are opaque in the visible part of the spectrum.

Take skin for example. Near IR can penetrate skin, which means it can be used to look at things just below the surface of our bodies, such as blood vessels. The article, “Body Vision: IR Applications and Lens Selection in Biomedical Settings,” (BioPhotonics, October 2010,) gives some examples.

Now I appreciate that most of my target audience deal with industrial applications of machine vision, so you may be wondering about the relevance of biomedical imaging to your work. Well here’s the point: IR light can be used to see through things that appear opaque to our eyes, and conversely, some materials that look transparent to us are opaque to IR.

Depending on what you’re doing, this can be either a blessing or a curse. Plastics, for instance, have some IR transmissibility, so a strong backlight might highlight internal features that can’t be seen otherwise. On the other hand, you may find that longer wavelengths – perhaps even just deep reds at 680nm – show some penetration, which will complicate surface inspection. (Metals of course are immune to these problems. Well I think they are. Try it for yourself.)

Bottom line? Next time you’re looking for a way to light that hard-to-inspect part, give IR a try. Or keep well away from IR. Either way, you might be surprised by the result.

Machine vision in the infra-red

I thought this was an interesting product announcement. Flir have been steadily building their presence in IR imaging, but historically they seem to have been aiming for the scientific, security, and maintenance markets. With the launch of the A615 – a VGA resolution GigE IR camera, they seem to be turning their attention to machine vision applications, particularly those in the food industry.

There are a couple of companies in the IR-MV space already – JAI and Sensors Unlimited – so this would seem to great a little more competition. That has to be a good thing, because one of the barriers to greater use of IR is price. Some of those cameras can be seriously expensive.

Look beyond the visible

Most machine vision systems are developed using visible light. That makes sense because most of the time we want to detect things that we can see with the naked eye. But interesting things happen when you go out into the infrared region of the spectrum.

Quick primer: visible light spans wavelengths from about 400nm up to 750nm. That’s from blue to red. As wavelengths get longer than 750nm we’re into the “infrared” region, which goes all the way up 1mm; the so-called “millimeter waves”. Now that’s a big range, so to make life easier scientists have divided the IR spectrum into three regions. There seems to be some dispute about the precise definitions, but in crude terms there’s Short Wave Infrared (SWIR,) which goes up to about 3,000nm, Mid-IR which covers the 3 to 30μm range, and Far-IR which is beyond 30μm.

The silicon sensors we use in our conventional machine vision cameras can’t detect anything past 1μm, but InGas detectors can go up to 3μm. So what can you see with InGas that you can’t see with a regular silicon CCD?

Well Sensors Unlimited Inc, part of Goodrich Corporation, have built a whole business around SWIR imaging, so if you visit their web site you’ll find plenty of examples. Some of the most interesting relate to the inspection of photovoltaic cells. For a short summary you light also like to read the article and watch the video, posted on Laser Focus World, (September 8, 2009.)

Now, a pop quiz: can you name one of the founders of Sensors Unlimited, and for a bonus point, what is his biggest claim to fame?