Qualitative classification of milled rice grains using computer vision and metaheuristic techniques

• Published in: Journal of food science and technology Vol. 53; no. 1; pp. 118 - 131
• Main Authors: Zareiforoush, Hemad, Minaei, Saeid, Alizadeh, Mohammad Reza, Banakar, Ahmad
• Format: Journal Article
• Language: English
• Published: New Delhi Springer India 01.01.2016; Springer Nature B.V
• Subjects: Accuracy; Algorithms; Analysis; Artificial intelligence; Automation; Chemistry; Chemistry and Materials Science; Chemistry/Food Science; Classification; Feature selection; Food; Food processing industry; Food Science; Grain; Methods; Morphology; Neural networks; Nutrition; Original; Original Article; Rice; Studies; Support vector machines; Topology; Vision systems; Computer vision; Metaheuristic techniques; Classification; Rice
• ISSN: 0022-1155; 0975-8402
• DOI: 10.1007/s13197-015-1947-4

Qualitative grading of milled rice grains was carried out in this study using a machine vision system combined with some metaheuristic classification approaches. Images of four different classes of milled rice including Low-processed sound grains (LPS), Low-processed broken grains (LPB), High-processed sound grains (HPS), and High-processed broken grains (HPB), representing quality grades of the product, were acquired using a computer vision system. Four different metaheuristic classification techniques including artificial neural networks, support vector machines, decision trees and Bayesian Networks were utilized to classify milled rice samples. Results of validation process indicated that artificial neural network with 12-5*4 topology had the highest classification accuracy (98.72 %). Next, support vector machine with Universal Pearson VII kernel function (98.48 %), decision tree with REP algorithm (97.50 %), and Bayesian Network with Hill Climber search algorithm (96.89 %) had the higher accuracy, respectively. Results presented in this paper can be utilized for developing an efficient system for fully automated classification and sorting of milled rice grains.