Discriminating robot system for carnation seedling with fuzzy logic

Abstract A carnation seedling is an agricultural product and its shape is very various. Therefore, it is difficult to discriminate this object by using the industrial method of some fixed pattern matching. We proposed the evaluation and the judgement of this seedling based on the subjective knowledge of a connoisseur, by using the image processing of the top and the side view and the fuzzy calculation such as fuzzy matching and fuzzy integral. In an online test with 500 samples, a good judgement rate of 97 % was achieved for a carnation seedling. This technique, mentioned here, will make it possible to evaluate and judge other agricultural product.