<![CDATA[Tomato plants as important horticultural commodities in Indonesia are vulnerable to various diseases that are difficult to distinguish visually, causing treatment delays and economic losses for farmers. This research develops an Android application for cloud-based tomato plant disease detection using Model-View-ViewModel (MVVM) architecture with Jetpack Compose framework. Extreme Programming (XP) methodology is applied through four main stages: planning, design, coding, and testing iteratively to ensure responsive development toward system requirements. The system integrates MobileNetV2 machine learning model deployed on Azure App Service with Firebase Authentication and Firestore for authentication and detection history storage. Dataset containing 16,012 tomato leaf images with 10 disease classes is used to train the model achieving 91.74% accuracy. Cloud-hybrid architecture enables detection processes to be performed on server without burdening mobile devices with average response time of 3 seconds. Black Box Testing shows all application features function according to specifications with 100% success rate. Research results prove that cloud computing integration with mobile application development can provide practical and scalable solutions to help farmers perform tomato plant disease detection quickly and accurately]]>
Copyrights © 2025
