Automated Chokun Orange Maturity Sorting by Color Grading

Abstract

An image processing technique is developed in order to guide Chokun orange maturity sorting. The objective of this research is to assess the fruit maturity by color grading. The process is divided into two major steps, the training step and the testing step. In the training step, images of 90 Chokun oranges of three different degrees graded by an experienced farmer are collected by a color digital camera under the normal illumination conditions with white fluorescent lamps. Then, the original RGB (Red, Green and Blue) color image of an orange is transformed into an ISH (Intensity, Saturation and Hue) image. From the hue images, the hue colors are analyzed and then used to form decision rules. A classifier is implemented using these decision rules. In the testing step, the degree of maturity of 50 Chokun orange samples is tested. The Chokun orange maturity sorting is done by using the classifier obtained from the training step. The experimental results show that the method of grading the Chokun oranges could be a feasible alternative with a success rate of about 98 %. Keywords: Chokun orange maturity sorting, decision rule, classifier, image processing, ISH color model