Cultivar Discrimination: CNN and Genetic Algorithms for Classifying Different Varieties of Ginger Plants

This study brings a new way to sort types of ginger in Indian farming. Using a Convolutional Neural Network (CNN) with a Genetic Algorithm (GA), the study tries to tell apart between Maran and Sunti. It uses $\mathbf{1 2 0 0}$ pics of ginger from different growth levels in various weather conditions. It is made better by GA to optimize it, this ensures it can pick up on little differences between ginger types well. By splitting the data into learning, checking, and testing parts, the model shows great skill by getting $93.28 \%$ right in identifying different types of ginger plants with high accuracy. Measures of how well a model works, like accuracy, recall, and F1-score show that it is solid and trustworthy for tasks where you need to classify things. This research is a major step forward in farming technology. It will make checking plants easier and improve ways to grow crops, leading to sustainable growth for India’s ginger industry.