Ensemble Classifier for Plant Disease Detection

Image processing refers to the technique for the extraction of digital information from images. These image processing techniques have been applied to various provinces including medical field, remote sensing, robot vision, pattern recognition, video processing, color processing, and so on in order to solve real world problems. These approaches have also been found out to be effective in the domain of Agriculture for the prediction of crop yield, leaf disease detection, fruit disease detection, vegetable quality evaluation, etc. Due to the immoderate usage of fertilizers and insecticides, the quality of plants and crops gets degraded to huge extent; and the plant diseases are generally detected by experts through naked eyes. In this paper, the authors proposed an ensemble model based on Random Forest and K-Nearest Neighbor (KNN) for the detection of plant diseases from the leaves. The experimentation has been carried out on benchmark dataset of Plant Disease Images. This dataset consists of 1000 images of Healthy Leaves and three plant diseases i.e. Brown Rust, Early Blight and Late Blight. The proposed approach works in four steps: Segmentation by K-Means Clustering, Feature Extraction by GLCM, Feature Selection by Random Forest, and Classification of plant disease by KNN. The performance of the proposed approach has also been evaluated with traditional support vector machine (SVM) on the basis of three measures such as accuracy, precision and recall. The experimental results revealed that the proposed approach outperformed SVM.