Assessing Paddy Field Health Using High-Resolution Multi-Spectral Camera in Subang Indonesia to Implement Precision Agriculture 4.0

Abstract

Increasing yield and improving the sustainability of agriculture are needed through data-driven precision agriculture technology. Multispectral sensor measures the light reflected from the plant canopy and generates a reflectance value that indicates the difference between healthy and stressed plants. We use precision agriculture to assess paddy field health in Subang Regency, West Java, Indonesia, especially early detection of bacterial leaf blight and stem borer disease. The condition of the study area has a relatively flat topography, with the Grendel glutinous paddy variety, and has a late vegetative to early generative phase. The data was gathered using an unmanned aerial vehicle (UAV) fitted with a MicaSense Red Edge-P sensor to obtain multispectral images. We employed Normalized Difference Red Edge Index (NDRE) to create a health model and validated using 70 field sampling locations. As a result, accuracy detection was 90% hence the use of UAV technology can be used for early disease detection. We classified the paddy field as healthy and unhealthy in percentage through spectral analysis and data from validation. This research has proven that we could detect disease in paddy fields faster and more efficiently. Utilizing UAVs for precision agriculture will prevent crop failure and increase food security in Subang and Indonesia. However, the limitation of this study is weather conditions and should consider plant varieties.