New algorithms used in the phases of automatic classification of products

The new automatic classification devices for products based on machine vision techniques value the shape and color parameters so as to suitably assess the latter. These parameters may also be turned to good account by the machine vision techniques that proved to be applicable in several domains. One can especially apply these techniques in the inspection and analysis systems of industrial products where the parameters vary according to very low limits. Taking into account these facts one has tried to develop a classification system of vegetal products based on machine vision techniques, as well as on decisive algorithms belonging to the artificial intelligence: neural networks and fuzzy algorithms. The phases of the classification process of vegetal products together with a part of the soft applications developed as well as some of the results obtained are presented in details. In plus, both the experimental methodology for classifying vegetal products and some original algorithms are presented in this paper. A new original program for algorithms implementation was developed and used during the tests.