Fig is one of the horticultural products in Iran which requires sorting at the postharvest stage in order to present to the market. In Iran, Fig is grading in two ways: by professional workers based on their being open-mouthed and by mechanical machines based on their diameters. This paper presents a new method based on artificial neural networks and vision machine for sorting of Figs. In this method, textural properties of co-occurrence matrix were extracted by using machine vision. Then, the figs were graded in three classes by artificial neural networks. The sorting accuracy for all classes was 89.5%.
