Disease Detection of Asian Rice (Oryza Sativa) in the Philippines Using Image Processing

In the Philippines, rice is one of the most important food consumed for almost every meal and so, it is essential to meet the demands in rice production. However, local storms contribute to the susceptibility of rice plants to diseases debilitating the ability of farms to produce large amount of high-quality rice. Identification of diseases are done manually by farmers based on experience or an expert's advice. This paper proposes an automated detection of diseases using image processing. In the segmentation technique, the images are divided into two sets where green pixels are masked with 1) blue and 2) black pixels, along with the other features, and are then fed separately into SVM and Random Forest to compare their performance. Results show that blue pixels and SVM as its classifier have yielded better outcome having 82.41% as compared to the other model.