Machine vision systems create images by analyzing the reflected light from an object, not by analyzing the object itself. Therefore it is critical to understand how light will reflect from a production part that one is inspecting. To understand this, one must understand the material that the object is made from, its size, shape, and finish. The object’s color is also important, even if the inspection system is monochromatic (a red part will reflect differently from a pink part). All of these characteristics affect how light is reflected, and may pose challenges in creating a good image. Finding the right light source can take a seemingly difficult application and simplify it.
Lighting can also affect how easy it is for computer analysis to produce a workable image. A good image is the result of proper focus and illumination. Images with poor contrast and uneven illumination require much more effort for the imaging computer, which ultimately increases the processing time, and can increase the number of incorrect image analyses. Generally, high contrast images are the easiest to process. Consider the image below (Figure 1). A good image processing system is capable of recognizing the plus signs in both images, however the repeatability of recognizing the top image will be lower due to the low contrast of this image compared to the one on the bottom. If illumination is properly controlled, the images should be handled by the machine vision system easily and with a high degree of reliability/repeatability.