Smart-Farming based Prevention System for Tomato Leaf Curl Disease

To monitor real time conditions of agricultural products like tomatoes, sensor-based monitoring can be implemented using machine learning techniques. AI based management can provide optimal environmental parameters to ensure the safety of crop production in a controlled environment. This paper presents an efficient methodology that utilize internet of thing (IoT) deployment for data collection, and a decision support system (DSS) is developed for decision making based on deep learning algorithm. The real time data (temperature, humidity, soil moisture) are collected from the tomato farms through IoT by deploying sensor nodes in a controlled environment. The DSS can assist farmers in management and automatically detect leaf curl disease in tomato plants. The real-time generated datasets are used in this research to train three separate models. The Random Forest model achieved the highest accuracy of 99.85% on the same dataset while SVM model has around 98.73% accuracy, and Decision Tree model obtained 98.57% accuracy. So, the Random Forest model is adopted and deployed for the early- stage identification of tomato and cucumber curl leaf disease.