A Quantitative Study of Illumination Techniques for Machine Vision Based Inspection

In this paper, three basic lighting geometries are compared quantitatively in an inspection task that checks for the presence of J-clips on an aluminum carrier. Two independent LabVIEW® machine vision algorithms were used to evaluate backlight, bright field and dark field illumination on their ability to minimize variations within a pass (clip present) or fail (clip absent) sample set, as well as maximize the separation between sample sets. Results showed that there were clear differences in performance with the different lighting geometries, with over a 30% change in performance. Although it is widely acknowledged that the choice of lighting is not a trivial exercise for machine vision systems, this paper provides a case study of the quantitative performance of different lighting geometries.