Ant Colony Optimization Algorithm for Disease Detection in Maize Leaf using Machine Learning Techniques

Abstract

Plant diseases have affected the productivity of food in recent years. Because of this productivity loss, humans and animals are also affected, but the whole biodiversity would be affected. So, we should take preventive measures to stop this food destruction. Both humans and animals largely consume the maize crop. Due to some factors, Maize leaf is easily affected by some fungal or other diseases. Farmers could not find out the leaf diseases at the early stages. They need some advanced methods to detect these types of diseases. Early detection of leaf disease helps farmers to increase the Maize yield. In the proposed algorithm, we have used five supervised machine learning algorithms such as K-nearest neighbor (KNN), naive bayes (NB), support vector machine (SVM), random forest (RF), logistic regression (LR) and then these machine learning models have been implemented with the Ant colony optimization (ACO) Algorithm for optimizing the accuracy of disease detection in Maize Leaf. For leaf classification, color and texture features are extracted from an input dataset. Features of a leaf can be described by Hu moments, Haralick texture, and color histogram. After performing all Machine learning classifiers, we have analyzed that Random Forest with Ant Colony Optimization Algorithm gives the highest accuracy of 99.4% for disease detection in Maize Leaf.