An Automated Plant Leaf Diseases Classification using AKMC and AKNN Machine Learning Techniques

Main Article Content

Mrs. R. Dhivya, Dr. N. Shanmugapriya


Agriculture is the largest determinant of a country’s economic development. Approximately 70 % of the Indian populace is reliant on farming. Caused by a multitude of climatic circumstances, plants become susceptible to severe disease outbreaks. Those diseases begin with the plant leaves and then spread to the rest of the plant, affecting both the quantity and quality of the crops that may be grown. A person's eye can't distinguish and diagnose the disease on each crop in the region because of the enormous quantity of plants. As a result, it's critical to properly identify the individual plant to prevent the illness from spreading. As a result of this research, we have developed a system for detecting and diagnosing plant leaf diseases using "Machine Learning (ML)" techniques, and then suggest the type of disease to cure on time. Our strategy is geared toward raising agricultural yields by focusing on agricultural production. In general, most methods employ a set of fundamental procedures, including such as data acquisition, preprocessing, segmentation, extraction of features, selection of optimal features, and classifications. In this research, we have proposed an "Advanced version of K-Means Clustering (AKMC)" approach for segmenting leaves image and an advanced version of the "K-Nearest Neighborhood (AKNN)" approach for classifying leaf diseases. The AKNN classification model has been used to recognize "Alternaria Alternata", "Anthracnose", "Bacterial Blight", "Cercospora Leaf Spot", and "Healthy Leaves" in the leaf image. Using a nonparametric procedure, this classifier can make classifications. Data for this research is obtained from the "Manu's Disease data set" from openly available resources. The programming tool "MATLAB R2017" with a system configuration "Intel Core i7 CPU with a 64-bit Operating System" was used to evaluate the simulations. The experimental parameters such as "Accuracy", "Recall (Sensitivity)", "Precision", and "F-measure" are compared and tabulated for the proposed "AKNN" and the existing "Bacterial Foraging Optimization based Radial Basis Function Neural Network (BRBFNN)". The performance of the proposed method surpasses the existing method according to the results.

Article Details