Use of machine learning for monitoring the growth stages of an agricultural crop informed by derived vegetation indices and spatial and temporal soil analyses Shara Ahmed, Nabanita Basu, Catherine E. Nicholson, Simon R. Rutter, John R. Marshall, Justin J. Perry and John R Dean Northumbria University, UK Oats ( Avena sativa L .) is Europe’s fifth largest crop and the sixth most grown cereal worldwide. Further, oat production in the UK reached 825 thousand tonnes in 2022, making it one of the most significant cereal crops grown in the UK. Therefore, to increase the crop yield of oats in response to this high demand, effective crop management strategies are required. Hence, the current study utilised an unmanned aerial vehicle (UAV) with multispectral image sensors to predict winter oats yield by the spectral indices of the normalised difference vegetation index (NDVI) and the chlorophyll index green (CI green) across three different growth stages of the oats crop. In addition, ground truth data relating to the actual crop yield, as well as soil health indicators and crop quality were collected. A hierarchical multinomial logistic regression machine learning model was incorporated to determine if the oats yield could be predicted using the soil health indicators and crop quality. The outcome of the machine learning approach provides a proportional range of soil nutrient levels and crop quality that farmers can use to anticipate the final oat grain yield prior to harvesting. The findings of this research study will be particularly valuable within a Precision Agriculture management strategy to anticipate soil interventions that lead to an enhanced oat crop yield.
P06
© The Author(s), 2023
Made with FlippingBook Learn more on our blog