Traditional optical sorting and grading systems use classic machine vision to look for features in the product image. These are known or expected features, such as the tip of a carrot or the flat bit of the stem on a leek when it transitions to the roots. Most of the time these features conform to a pattern or shape that is expected. With an organic object such as a vegetable, there is never a fixed size, shape or colour of anything, and this presents a problem to classic camera systems that only rely on known shapes or patterns because the variation means that there will always be a high percentage of unknowns.
In an automation system that relies on classic machine vision, this can translate to a relatively poor yield output from the machine.
For more information download our case study