Vision systems for quality control

To successfully implement a vision system into industrial quality control, several resolution aspects must be met. First is temporal resolution, being able to capture the image in a way which isolates features in time. The rate of image capture in a vision system can not be a bottleneck to upstream production processes. Another is spatial resolution, the vision system must be able to see the minimum detectable feature on every sample imaged. The third is gray scale resolution, the number of steps between the system noise floor, and maximum light the system can absorb. The fourth is spectral resolution, or color. The way to meet these criteria does not typically require a significant amount of technical sophistication. This paper discuss some of the more significant criteria needed to be resolved in order to meet the temporal and spatial resolution aspects in a vision system, and the cost tradeoffs involved in meeting the requirements.<>