Studies on Image Recognition of Fruit Vegetables (Part 1)

Extraction of Feature Parameters on Binary Image of Fruit Vegetables

Abstract

In this report, feature parameters related to the complication and the direction such as complexity, fractal dimension, Feret's diameter ratio, moment and others were extracted on binary image of standing fruit vegetables in order to establish a relatively simple method of image recognition. The possibility of the shape recognition was investigated.
From the results, it was found that these feature parameters could express the plants from the various viewpoints, though the values scattered because of the complication and the manifold of the plant. It was considered that these parameters could be effectual for the image recognition if these were used in proper combination, instead of using independently.