Assessment of Damage Due to Grain Discoloration Disease in Paddy Crop Using High-Resolution UAV Imagery

Assessment of damage area and yield due to disease in paddy crop are required to be more precise, quick, and inexpensive. Traditional and satellite approaches of damage assessment are inaccurate and time-consuming in quantification for small agricultural fields. Unmanned Aerial Vehicle (UAV)-based remote sensing is considered to be a promising tool to determine the damage of target crops on a large scale with high spatio-temporal resolution. The aim of this study is to assess the damages due to disease of paddy (MTU-1010) crop using UAV-based multispectral imageries. UAV-based multispectral imageries are obtained for Kharif 2018 and 2019. Eleven vegetation indices are calculated using multispectral images to assess damages. A good correlation (R²) is obtained between Normalised Difference Vegetation Index (NDVI) and yield (0.765 and 0.647 for Kharif 2018 and 2019, respectively). Results obtained from Kharif 2018 are validated with Kharif 2019. The overall estimated yields for Kharif 2018 are within 10% error (Root mean square error (RMSE) = 0.042 kg m^− 2 and Relative RMS error = 0.094), whereas 60% of estimated yields for Kharif 2019 are within 10% error (RMSE = 0.063 kg m^− 2 and Relative RMS error = 0.15). For both years, NDVI, an indicator of grain discoloration disease in paddy, ranging from 0.511 to 0.715, 0.716 to 0.745, and > 0.745, are used to describe crop conditions for diseased, moderately healthy, and healthy crops, respectively. Thus, multispectral imageries acquired by UAV are useful to detect and assess the damage in terms of yield and areal extent of paddy quickly and efficiently.