Study on Grade Judgment of Fruit Vegetables Using Machine Vision (Part 3). Judgment for Shape of Green Pepper Using Neural Network.

Abstract

A new machine vision system was developed to grade fruit vegetables. The system has learning ability and can therefore grade various fruit vegetables. The system was previously tested for shape grading of strawberries. This report describes the use of the new machine vision system to judge the shape of green peppers. Because of the more complicated shape of green peppers, the application software of the system was improved for judging accuracy, including new extraction methods for shape features and compound grading methods using binary logic and neural networks. The grading processes using the developed 3 application software are also described. Test results showed that the new machine vision system with binary logic and a multiple-layer neural network has great potential for automatically sorting green peppers.