Shape Classification of Asparagus Using Fuzzy Logic and Neural Networks

Abstract

The grade of agricultural products in Japan is in many cases based on the degree of bend in the product, and the degree of curvature is especially reflected in the price of asparagus and cucumbers. Because this property has many aspects, machine classification is not easily achievable. However, there is an increasing demand for automated grading of seasonal agricultural products. In order to classify shapes of asparagus automatically, we applied the fuzzy logic to determine the degree of curvature and a neural network to classify shapes of this product. Classification by this system has a 93% coincidence with results from visual classification by experts.