Detection and classification of fungal disease with Radon transform and support vector machine affected on cereals

This paper describes Radon transform and SVM-based recognition and classification of visual symptoms affected by fungal disease. Algorithms are developed to acquire and process colour images of fungal disease affected on cereals like wheat, maize and jowar. Different types of fungal disease symptoms namely, leaf blight, leaf spot, powdery mildew, leaf rust, smut are considered for the study. The developed methodology consists of two phases. In the first phase, Radon transformation and projection algorithm is used to project patches (affected area) on the surface of cereal and detect whether the cereal is fungal affected or normal. In the second phase, fungal affected symptoms are classified using support vector machine (SVM) classifier. The fungal affected regions are segmented using k-means segmentation. Colour and shape features are extracted from affected regions and then used as inputs to SVM classifier. Classification accuracies of over 80.83% are obtained using colour features, 85% are obtained using shape features and 90.83% are obtained using combined colour and shape features.