Classification of Rubber Tree Leaf Diseases Using Multilayer Perceptron Neural Network

Abstract

This paper presents about classification of rubber tree leaf diseases through automation and utilizing primary RGB color model. Several rubber tree leaf diseases are been studied for digital RGB color extraction where three sets of rubber tree leaf diseases image are digitally captured under standard and control environment. The identified regions of interest (ROI) of these diseases images are then processed to quantify the normalized indices from the RGB color distribution. This system involved the process of image classification by using artificial neural network where 600 samples were used as training while another 200 samples were for testing. The optimized ANN model in this work has two method which based only on the dominant pixel RGB (mean) and applying principle component analysis (PCA) on the pixel gradation values of each image. The optimized model was evaluated and validated through analysis of the performance indicators. Findings in this work have shown that both models have produced about 70% in diagnostic accuracy with more than 80% achievement for sensitivity. However, model with the applied PCA has lower network size.