4.6 Article

An accurate method for predicting spatial variability of maize yield from UAV-based plant height estimation: a tool for monitoring agronomic field experiments

期刊

PRECISION AGRICULTURE
卷 22, 期 3, 页码 897-921

出版社

SPRINGER
DOI: 10.1007/s11119-020-09764-w

关键词

Unmanned aerial vehicle (UAV); Crop surface model (CSM); Yield prediction; Photogrammetry; Biomass estimation

资金

  1. scientific direction of AgroParisTech
  2. French national observatory networks SOERE PRO part of the AnaEE-France project of the French Investments for the Future (Investissements d'Avenir) program [ANR-11-INBS-0001]

向作者/读者索取更多资源

Estimating aboveground biomass and yield in agronomic field experiments using UAVs showed a 15% improvement in accuracy compared to manual estimation, offering a new feasible solution for monitoring crop growth.
Estimating aboveground biomass is important for monitoring crop growth in agronomic field experiments. Often this estimation is done manually, destructively (mowing) or not (counting) on a relatively limited number of sub-plots within an experiment. In the presence of spatial heterogeneity in experiment fields, sensors developed for precision agriculture, have shown great potential to automate this estimation efficiently and provide a spatially continuous measurement over an entire plot. This study investigated the suitability of using an unmanned aerial vehicle (UAV) for biomass and yield estimations in an agronomic field experiment. The main objectives of this work were to compare the estimates made from manual field sampling with those made from UAV data and finally to calculate the improvement that can be expected from the use of UAVs. A 6-ha maize field was studied, with plot treatments for the study of the exogenous organic matter (EOM) amendment effect on crop development. 3D surface models were created from high resolution UAV RGB imagery, before crop emergence and during crop development. The difference between both surface models resulted in crop height which was evaluated against 38 reference points with an R-2 of 0.9 and prediction error of 0.16 m. Regression models were used to predict above-ground biomass and grain yield (fresh or dry). Dried grain yield prediction with a generalized additive model gave an error of 0.8 t ha(-1) calculated on 100 in-field validation measurements, corresponding to a relative error of 14.77%. UAV-based yield estimates from dry biomass were 15% more accurate than manual yield estimation.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.6
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据