Showing posts with label polarization. Show all posts
Showing posts with label polarization. Show all posts

Tuesday, August 9, 2011

Understanding polarization


From time to time, as you browse the literature put out by makers of machine vision lighting, you’ll come across references to polarization. As a well-schooled engineer you’ll know that a polarizing filter is the optical equivalent of a venetian blind, only allowing light whose electric field oscillates in a particular plane to pass, but do you really understand the how and why?

I believe that making use of a technology or physical effect depends on understanding the basic principles, so today I’d like to share a few links to pages that deal with polarization.

Let’s start by talking about what polarization is. Essentially, it’s a result of reflection from a surface. Yes, angle of incidence equals angle of reflection, but the oscillating electric field makes things a bit more complicated than that. To put it crudely, when light reflects off a surface towards a camera, the light entering the lens all has the same polarization. A great explanation is provided by Professor Orzel of Union College, Schenectady, NY under the heading of “Basic Concepts: Polarization of Light”.

Then, when you’ve understood that article, take a look at his follow-up, “Why Do Polarized Sunglasses Work?” which actually gets into the nitty-gritty of what causes polarization.

From there it’s time to think about how to take advantage of polarization and that’s where I’m going to send you to a short article on the Edmund Optics website called, “Successful Light Polarization Techniques”. This has some photos showing how polarizing filters can clean up an image, making it more suitable for processing.

One point that I didn’t see mentioned anywhere though concerns aperture and exposure settings. Only allowing light of a certain polarization to reach the sensor means that the total number of incident photons is reduced. In other words, the image will get darker. To compensate for this you’ll have to open the lens aperture, increase the exposure time, or just throw more light at your subject.

Now, that’s my brief polarization primer over; let’s turn it over to you my dear reader. How about sharing some examples of clever or intelligent uses of polarization?

Tuesday, July 20, 2010

Useful backlighting tip

If you’ve ever tried backlighting a polished metal cylinder – like a roller bearing, for example – you’ll know it’s not as easy as much of the machine vision literature would imply. The problem is that if your backlight is larger than the diameter of the cylinder you will inevitably see reflections from the sides of the pin. This makes it difficult to ascertain the actual edge and as a result, diameter measurements will be unreliable.

One solution is to use a telecentric or collimated backlight. Edmunds make a range, However, they are rather expensive and are only available in smaller diameters. But help is at hand!

Spencer Luster of Light Works recently published an email tip suggesting an alternative technique, one that I’d never heard of. This involves using two circular polarizers either side of the cylinder you’re trying to gauge. I don’t really understand all the physics of polarization, but judging by the images Spencer offers, it seems to do the trick.

I couldn’t find the paper on the Light Works web site, so if you want to learn more I suggest you use the “Contact” button and email a request. I’m sure Spencer will be very pleased to hear from you.