Title: An intelligent rice quality classifier
Authors: L.A.I. Pabamalie; H.L. Premaratne
Millennium Information Technologies Ltd., No. 1, Millennium Drive, Malabe, Sri Lanka.
University of Colombo School of Computing, No. 35, Reid Avenue, Colombo 7, Sri Lanka
Abstract: Amidst the laws enforced on consumer protection, the release of inferior quality food to the consumer market takes place, particularly more often in developing countries. At present, in Asian countries where rice is a principal meal, quality checks are carried out through visual inspection methods. Therefore, an effective quality testing mechanism is necessary to ensure the quality of food meet required standards. Although some research has been reported on the classification of paddy seeds, no published work is found on the classification of milled rice. This research, with the aim of implementing a hand-held device for quality testing of rice, focuses on producing a classification system based on neural network and image processing concepts. Thirty one texture and colour features are extracted from rice images for discriminate analysis. Tests on the system for the training and test sets show the accuracy in between 94% to 68% for the four grades.
Keywords: rice quality; classification; neural networks; image processing; grey-level co-occurrence matrix; GLCM; regression analysis; root mean square error; RMSE; mean absolute error; MAE; food quality; developing countries; milled rice; quality testing; food standards.
Int. J. of Internet Technology and Secured Transactions, 2011 Vol.3, No.4, pp.386 - 406
Available online: 17 Oct 2011