Classifying Quality Grading and Size of Allium Cepa Using Digital Image Processing Algorithms

Abstract

This study introduced the system that can classify the size, color, and texture of the onion based on Philippine National Standard (PNS) using image processing methods such as Gaussian Blur, Canny Edge, Erosion, Dilation, Contouring, Color Masking, Pixel per metric ratio, RGB, and Blob. The Blob algorithm can detect the texture of the onion according to its grade; Grade A(Grade 1), Grade B (Grade 2), Grade C (Combination of Grade 1 and Grade 2), and Reject. The researcher conducted series of experimental testing to compare the quality grading using manual and automatic classification of bulb onion. The statistical method used for this study is T-test to compare the manual sorting and automated sorting. The t-values for red onion based on local standards is 0.1284, while the yellow onion is 0.0178. The results are less than the critical value of 2.01. It means that the null hypothesis is accepted. Over all, the system can determine the size, color, and texture bulbs of onion using image algorithms.