Olive Leaf Disease Detection using Improvised Machine Learning Techniques

Abstract

Plants are integral to human life, and so, plant health is important. Regularly monitoring of plant health and plant disease detections are important in property agriculture. In agriculture, the use of image processing techniques run by computers in solving agricultural problems is increasingly common, particularly in the classification and identification of crop disease. Such usage could preserve the technical and commercial well-being of agriculture. This study demonstrated the application of support vector machine and image processing-enabled approach to detect and classify Olive leaf disease. It comprises seven steps that begin with a presentation of a digital color picture of a sickly leaf, followed by the step of image denoising using mean function, image enhancing using CLAHE method, image segmentation using fuzzy C Means algorithm, image feature extraction using PCA, and disease detection and classification using PSO SVM, BPNN, and random forest algorithms. The results showed high accuracy of the proposed PSO SVM in Olive leaf disease classification and detection.