Monday, May 18, 2009

Dealing with laser speckle

If you’ve ever tried using a laser line generator you’ll have noticed the “speckle” in the line. This isn’t the place to explain why speckle occurs - I’m focused on solutions here, not basic physics. Suffice to say that the laser does not provide a homogeneous line of light. To make matters worse, if the object you’re illuminating moves, the bright spots of speckle will also dance about in a somewhat random pattern.

This is a problem if you need to find the center of the laser line, something you might want to do if you’re trying to determine relative heights. So what options do you have?

I suggest the following:
  1. Let the object you’re measuring move and use a long exposure. This will average out the speckle.
  2. Open up the iris and allow a limited amount of saturation to occur.
  3. Perform some morphology on the images.
In short, lasers produce a nice line, but look closely to see if the raw image is good enough. It might still be necessary to do some image processing.

No comments: