Illumination for Machine Vision
We define machine vision as a field of 'artificial intelligence' that, by using the suitable techniques, allows to obtain, process and analysis any type of information given through digital images.
Illumination is the most important part of an imaging system. Choosing an inappropriate illumination could determine if a system works reliably or does not. Optimizing the lighting can often eliminate the need for costly and time consuming image manipulation.
Cameras capture light reflected from objects and are much less versatile than human eye. Optimizing lighting conditions we improve the detection and obtain a vision equal to or even better than the human eye. The aim of the illumination used in machine vision applications is to control how the object appears to the camera and this way, to highlight those aspects we are interested on. For instance, light reflected from a coin is not the same as light reflected from a sheet of white paper, so that they require different lighting techniques.
Factors affecting lighting
There are some requirements we must consider when determining which light is better for an application: How is the speed of inspection?, How is the field of vision to illuminate?, Is it monochrome or colored?, Which is the feature to highlight?, Which mechanical or physical requirements we must comply?
These requirements are directly related to the factors affecting lighting:
► Types of lighting (angle of illumination).
► Color of light. Wavelengths.
► Type of light: Direct, diffuse and ultra-diffuse.
There are so many factors that could determine the best lighting for each application, and is hard to establish predetermined rules for selecting the suitable light due to the fact of all those parameters mentioned before. For this reason, the best option is to ask our technical department for advice. They would effectively solve all your doubts.
