machine learning-based in-season nitrogen status prediction using unmanned aerial vehicle remote sensing

Objective the uniformity of fertilizer application across fields is a common practice, driven by local legislation or by expert opinion. however, this approach might lead to over-application of nitrogen in areas with poor yields. human health, ecological functions, biodiversity, climate change, and long-term stability are all adversely affected by the increasing release of reactive nitrogen into the environment that may result from the excessive use of fertilizers. the purpose of this work was to show that throughout the growth season, location-specific n proposals may be generated using non-invasive crop status monitoring that is built on remote sensing technologies (rst). this tracking system can accurately assess the position of crop n. materials and methods in this study, two frameworks—support vector machine (svm) and artificial neural networks (ann)—that rely solely on data collected from crop sensors, with the goal of improving our ability to predict crop n nutrition index (nni) and crop yield throughout the growing season were compared. this was performed by combining data from soil, weather, and cultivation with information from present detectors using random forest (rf). results through rst, a simple and low-cost tool known as a fixed-wing unmanned aerial vehicle (uav) may capture wavelength reflection images. this collection of images is invaluable to polystyrene nanoparticles (pnsp). applying ml techniques enhanced the nni estimate, as seen by the results.conclusions utilizing machine learning techniques presents a valuable opportunity to maximize the use of rst data, enabling more effective monitoring of agricultural production factors and directing pnsp strategies

machine learning-based in-season nitrogen status prediction using unmanned aerial vehicle remote sensing

objectivethe uniformity of fertilizer application across fields is a common practice, driven by local legislation or by expert opinion. however, this approach might lead to over-application of nitrogen in areas with poor yields. human health, ecological functions, biodiversity, climate change, and long-term stability are all adversely affected by the increasing release of reactive nitrogen into the environment that may result from the excessive use of fertilizers. the purpose of this work was to show that throughout the growth season, location-specific n proposals may be generated using non-invasive crop status monitoring that is built on remote sensing technologies (rst). this tracking system can accurately assess the position of crop n.materials and methodsin this study, two frameworks—support vector machine (svm) and artificial neural networks (ann)—that rely solely on data collected from crop sensors, with the goal of improving our ability to predict crop n nutrition index (nni) and crop yield throughout the growing season were compared. this was performed by combining data from soil, weather, and cultivation with information from present detectors using random forest (rf).resultsthrough rst, a simple and low-cost tool known as a fixed-wing unmanned aerial vehicle (uav) may capture wavelength reflection images. this collection of images is invaluable to polystyrene nanoparticles (pnsp). applying ml techniques enhanced the nni estimate, as seen by the results.conclusionsutilizing machine learning techniques presents a valuable opportunity to maximize the use of rst data, enabling more effective monitoring of agricultural production factors and directing pnsp strategies