Tugas Akhir
Design of an iot-based powdery mildew disease detection system on melon plants using a webcam and dht22 sensor / Jessica Budymas Manora Yollanda
Abstrak
Melon plants (Cucumis melo L.) are high-value horticultural commodities that are susceptible to powdery mildew disease. This disease is caused by the fungi Erysiphe cichoracearum and Sphaerotheca fuliginea which thrive in warm temperatures and high humidity. Powdery mildew infections result in reduced yield quality and quantity. This study aims to design an Internet of Things (IoT)-based detection system for powdery mildew disease that can monitor environmental conditions and visually detect disease symptoms using a webcam and DHT22 sensor. The method used is a quantitative approach involving the design of a system prototype consisting of a Raspberry Pi as the microcontroller a webcam for leaf image capture and a DHT22 sensor for temperature and humidity monitoring. The system sends real-time notifications to a Telegram application if environmental conditions or visual symptoms favorable to fungal growth are detected. Testing results show that the system is capable of accurately detecting changes in temperature and humidity. The webcam achieved a disease detection accuracy rate of over 90% under bright lighting conditions although accuracy decreased under very low light. The DHT22 sensor effectively monitored temperature and humidity as contributing factors to fungal growth. Notifications were delivered to Telegram within 1 ndash 2 seconds indicating a fast system response.