Artificial bee colony-based fuzzy c means (ABC-FCM) segmentation algorithm and dimensionality reduction for leaf disease detection in bioinformatics

Abstract

The rapid increase in human population has necessitated a corresponding increase in agricultural production. The advancements made in the arena of genomics and bioinformatics can open doors for this. As disease detection is a primary factor in enhancing agricultural production, the current research focuses on coming up with a sound plant leaf disease detection and identification procedure for large areas of crop production. Since dimensionality reduction plays a significant role in effectively detecting a plant leaf disease, this study achieves this by using singular value decomposition (SVD) technique. At first, the leaf images are preprocessed, and background subtraction is carried out using Gaussian mixture model. After this, the diseased area is segmented with the introduction of artificial bee colony-based fuzzy C means algorithm. Then, the detection rate is increased by applying SVD, which reduces the dimensions of multiple feature vectors. Finally, the leaf disease is detected with the help of multi-kernel with parallel deep learning classifier. The implementation of all these steps is done via MATLAB simulation environment. The evaluation of the results of the proposed leaf disease approach is done for metrics including accuracy, recall, precision and F-measure.