SMARF: Smart Farming Framework Based on Big Data, IoT and Deep Learning Model for Plant Disease Detection and Prevention

Basori, Ahmad Hoirul and Mansur, Andi Besse Firdausiah and Riskiawan, Hendra Yufit (2020) SMARF: Smart Farming Framework Based on Big Data, IoT and Deep Learning Model for Plant Disease Detection and Prevention. Applied Computing to Support Industry: Innovation and Technology, 1174. pp. 44-56. ISSN 1865-0929

Text (Hasil Turnitin) 13. SMARF__Smart_Farming_Framework_Based_on.pdf Download (2MB)

Abstract

Plant disease can become a serious threat toward food production and security since the demand for food increased significantly over the year. The big data and deep learning have been discussed and explored highly in recent years due to its capability to detect certain features in smart ways. Whilst, crop disease that attack leaves can be cured if farmer detects the early symptoms and avoid the spreading of the disease. This paper presents the capability of big data and deep learning to give predictive analytic toward the plant crop disease. Some features such as leaves, weather, soil and other landscapes condition are taken as an input for the system. Smart farming will utilize IoT technology on capturing the data and localize the position of the infected plant. The combination of computer vision and GPS technology will be able to pinpoint the disease location in efficient ways. The experimental result has shown that deep learning is superior compare to logistic regression with 72% accuracy of identification of an infected leaf. Of course, this result can be augmented further by involving the extra of the leaf. Overall, the Smart Farming framework is able to give a better solution for plant disease spreading prevention by early detection and localization of the disease. A farmer might get advantage by this notification, especially for wide-scale farming. The future work might involve real-time data from the drone or CCTV camera in the real farming field.

Actions (login required)

View Item