One of the newest trends in machine vision systems is the implementation of so-called "smart cameras." A smart camera combines the usual image sensor with a built-in processor, which allows inspections to be run directly on the camera, thereby eliminating a step in the process. Instead of simply capturing images, like a conventional camera, a smart camera can analyze those images, interpret the data, and deliver the results. Smart cameras are also designed to perform many advanced functions normal cameras can't such as edge detection, geometric pattern matching, optical character recognition, and 2-D barcode reading.
The hardest part of building a machine vision inspection system utilizing smart cameras is programming the application. For machine builders who aren't comfortable writing software, there are two basic approaches they can take. The first is to configure the inspection system using a commercially available, no-programming, menu-driven machine vision software package. The second option, for engineers who have experience with a graphical programming language, is to use NI LabVIEW RealTime with NI's Vision Development Module.
In a feature article in the March issue of Imaging Technology magazine, Matthew Slaughter of National Instruments explains some of the finer points of designing machine vision systems with smart cameras. Read the full article on page IIa of the March issue of Imaging Technology.