Klasifikasi Penyakit Tanaman Padi Mengunakan Metode Convolutional Neural Network Melalui Citra Daun (Multilayer Perceptron)

Abstract

Rice is a staple food so its production must be maintained. Disruption of rice production can be caused by disease disorders in rice plants. One of the diseases in rice plants is leaf disturbance. The types of diseases in rice plant leaves are very diverse, the types of diseases are well recognized by experts. Diseases of a plant will greatly affect the yield of the crop. If the disease is not treated immediately, the disease can damage crops and result in crop failure which will affect the economy. Therefore, the detection of plant diseases is very important in the process of plant care. Diseases that most often attack rice plants in Indonesia are Bacterial Leaf Blight, Brown Spot and Leaf Spot. Therefore, this study aims to classify rice diseases based on the image of the leaves that have been attacked by the disease using Deep Learning techniques with the Convolutional Neural Network (CNN) method. parameter scenario comparison dataset 90%, size 100x100 pixels, kernel 3x3, learning rate 0.01, Adam optimizer, epoch 150 and batch size 30.