For any gardener recognizing which plants they have in their plots or pots is a challenge especially if you're growing different varieties of the same species. For amateurs, there are apps available which crowd-source information. But for farmers, trying to grade the quality and even variety of their produce is a perennial problem. One variety of strawberry looks pretty much the same as another to the untrained eye and even long-term farmers may struggle, as do peppers, tomatoes, potatoes, and so on. So for a commercial enterprise being able to grade and classify their produce is essential if they're to get the best prices and avoid mistakes.
Traditionally, fruit and vegetables are graded manually, where pickers look at a piece of produce and place it in the appropriate container. There are automatic graders available and these use weighting and these use weighting and PLCs to sort the produce into the appropriate container. According to this website, grading machines work on the following principle "The automatic fruit and vegetable sorting machine adopts balance and lever principle using container weight and weight apparatus set weighting, achieving classification according to weight during the shift and measurement of moving. In the automatic transmission, it can grade all kinds of fruit and vegetables quickly and accurately." Such is the portability of these machines that they are often taken to the field where the produce is picked.
But, there is an obvious problem with this system. It can't spot the variety or see any spoilt vegetables. But now a solution may be at hand. Machine vision systems, so often used in manufacturing industries to spot defects and type of parts, can now be used in the field for the farming industry. According to this paper from 2008 and his one from 2011, machine vision systems can be used to compare the details of the fruit passing underneath a camera with the details on a database. For example, on the database are ideals of what the "perfect" dimensions, coloring, and shape. Depending on the results, the produce will then be sent to the appropriate "bin". This ides is already being adopted by producers as this video for grading dates demonstrates.