Agarwood Chips Grading Using Neural Network

  • Marini Nafi Politeknik Kota Bharu
  • Murni Rahim Politeknik Kota Bharu
  • Siti Khairunniza Bejo Universiti Putra Malaysia

Abstract

This paper presents the use of image processing and artificial neural network to determine the grade of the agarwood chips and thus provide an automated approach of the agarwood grading system. A backpropagation multilayer feedforward neural network has been used in this study with the inputs taken from the texture measurements and density. The relationship between texture properties and the price of agarwood was analyzed in order to select suitable input parameter to the neural network model. As a result, neural network architecture with three input parameters taken from textural properties and one input parameter taken from density, one hidden layer, seven number of neurons in hidden layer and three output layers has been developed. Neural network with algorithm of traincgp and transfer function of purelin and tansig give the best result of prediction with the percentage of accuracy of 60.58%.

Published
2017-11-16
How to Cite
NAFI, Marini; RAHIM, Murni; BEJO, Siti Khairunniza. Agarwood Chips Grading Using Neural Network. Politeknik & Kolej Komuniti Journal of Engineering and Technology, [S.l.], v. 2, n. 1, p. 31-47, nov. 2017. ISSN 0128-2883. Available at: <http://myjms.mohe.gov.my/index.php/PMJET/article/view/2936>. Date accessed: 17 aug. 2019.