首页|University of Nebraska-Lincoln Researchers Provide New Insights into Machine Lea rning (Integrating UAV hyperspectral data and radiative transfer model simulatio n to quantitatively estimate maize leaf and canopy nitrogen content)
University of Nebraska-Lincoln Researchers Provide New Insights into Machine Lea rning (Integrating UAV hyperspectral data and radiative transfer model simulatio n to quantitatively estimate maize leaf and canopy nitrogen content)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – Investigators discuss new findings in artificial intelligence. According to news reporting originating from Lincoln, N ebraska, by NewsRx correspondents, research stated, “Crop nitrogen (N) content r eflects crop nutrient status and plays an important role in precision nutrient m anagement. Accurate crop N content estimation from remote sensing has been well documented.” The news journalists obtained a quote from the research from University of Nebra ska-Lincoln: “However, the robustness (i.e., the ability of a model to perform c onsistently across various conditions) of these methods under varied soil condit ions or different growth stages has rarely been considered. We proposed a hybrid method that integrates in-situ measurements and the data simulated by a mechani stic model to improve the estimation of maize N content. In-situ data included h yperspectral images collected by Unmanned Aerial Vehicle (UAV), and leaf and can opy N content (LNC and CNC). A mechanistic radiative transfer model (PROSAIL-PRO ) was used to generate simulated data, i.e., canopy reflectance paired with targ et crop traits (i.e., LNC, CNC). We compared the performance from the hybrid met hod with a machine learning method (Gaussian Process Regression) and six differe nt vegetation indices (VIs) on four in-situ datasets collected at three study si tes from 2021 to 2022. Results show that the hybrid method consistently performe d the best for LNC estimation across four testing datasets (RRMSE ranging from 1 0.08% to 10.84%). For CNC estimation, the hybrid meth od had the best estimation results on two out of the four testing datasets and p erformed comparably to the best method on the other two datasets (RRMSE ranging from 13.89% to 25.21%). Next, we assessed the estimat ion robustness of the hybrid method, the machine learning, and the best-VI by co mparing the mean () and standard deviation (s) of RRMSE across diverse water and N treatments (condition #1) and different growth stages (condition #2). Among 16 total cases (two crop traits by four study sites by two conditions), the hybrid method had 11 cases of smallest and seven cases of s mallest s, outperforming the machine learning (0/16 for , 4/16 for s) and the be st-VI (5/16 for , 5/16 for s). These results underscore the greater robustness o f the hybrid method.”
University of Nebraska-LincolnLincolnNebraskaUnited StatesNorth and Central AmericaCyborgsEmerging Technolog iesMachine LearningNitrogen
NETL
NSTL
