Split screen

Back in the 1970’s, (yes, I am that old!) movie producers became very fond of the split screen technique. This allowed them to show both ends of a phone conversation, though some used it for more “artistic” effects too.

Well the split screen technique has advantages in machine vision. The conventional approach to capturing two views of an object is to use two cameras, two lenses, and two lights. But as you know, the objects we look at in machine vision seldom feature a convenient 16:9 aspect ratio, so many of the pixels in each image are wasted.

Split-screening means using a mirror or prism to let a single camera see the target from two directions. This can result in more efficient use of camera pixels as well as considerable cost savings.

A great example of this comes from National Instruments Alliance partner Coleman Technologies. A case study titled “High-Speed Seed Counting With LabVIEW and NI Vision” describes how a mirror was used to generate a second view for a single linescan camera. Not only did this save hardware costs, but, according to the author, it also simplified the image processing.

Incidentally, Coleman Technologies might sound familiar to regular readers. That’s because, back in October of 2011 I described an application of theirs that had been featured in Vision Systems Design magazine. The title of my articles was “Learning from others” and the application was inspection of dinnerware.