Betel Leaf Disease Detection Using Machine Learning

Abstract

This report discusses the design, development, and evaluation of a Betel Leaf Disease Detection System that aims to utilize machine learning and mobile technology to enhance agricultural practices. The main goal of the proposed system is to automate the process of finding diseases in betel leaf, which is an important crop in many areas. This will help farmers get help quickly and stop them from losing crops. The project will solve both farming and computing problems by making deployment easy, accurate, and efficient. The research commences with a literature review that delineates various machine learning models utilized for disease detection, including CNN, VGG16, MobileNetV2, and InceptionV3. Plant disease detection has come a long way in a short amount of time, but there is still a big need for crop-based datasets and real-time mobile apps, especially for betel leaf. To fill this gap, a system is created that combines machine learning with mobile deployment. This lets farmers detect diseases in real time and even from remote areas. We look at the model's performance by looking at its accuracy, precision, recall, and F1-score. We looked at the classification performance of models like VGG16, MobileNetV2, and InceptionV3. We found that MobileNet V2 has the best accuracy-to-complexity ratio for mobile deployment. Using confusion matrices to compare models shows the pros and cons of each model when it comes to classifying diseases. The last part of the project talks about how the system affects society, standardization, sustainability, and future work, such as improvements like treatment recommendations. This study paves the way for future advancements in machine learning within agricultural technology and presents an economical and comprehensive approach to enhance crop management.