Estimating Biomass and Nitrogen Amount of Barley and Grass Using UAV and Aircraft Based Spectral and Photogrammetric 3D Features

Abstract

Abstract:The timely estimation of crop biomass and nitrogen content is a crucial step in various tasks in precision agriculture, for example in fertilization optimization. Remote sensing using drones and aircrafts offers a feasible tool to carry out this task. Our objective was to develop and assess a methodology for crop biomass and nitrogen estimation, integrating spectral and 3D features that can be extracted using airborne miniaturized multispectral, hyperspectral and colour (RGB) cameras. We used the Random Forest (RF) as the estimator, and in addition Simple Linear Regression (SLR) was used to validate the consistency of the RF results. The method was assessed with empirical datasets captured of a barley field and a grass silage trial site using a hyperspectral camera based on the Fabry-Pérot interferometer (FPI) and a regular RGB camera onboard a drone and an aircraft. Agricultural reference measurements included fresh yield (FY), dry matter yield (DMY) and amount of nitrogen. In DMY estimation of barley, the Pearson Correlation Coefficient (PCC) and the normalized Root Mean Square Error (RMSE%) were at best 0.95% and 33.2%, respectively; and in the grass DMY estimation, the best results were 0.79% and 1.9%, respectively. In the nitrogen amount estimations of barley, the PCC and RMSE% were at best 0.97% and 21.6%, respectively. In the biomass estimation, the best results were obtained when integrating hyperspectral and 3D features, but the integration of RGB images and 3D features also provided results that were almost as good. In nitrogen content estimation, the hyperspectral camera gave the best results. We concluded that the integration of spectral and high spatial resolution 3D features and radiometric calibration was necessary to optimize the accuracy.

Canopy height measurements and non‐destructive biomass estimation of Lolium perenne swards using UAV imagery

“…We tested whether the combination of different types of data resulted in a better DMY prediction and investigated the influence of the methodology selected for the regression. As demonstrated by Bendig et al () and Brocks and Bareth () for barley, Yue et al () for winter wheat, Näsi et al () for a timothy and meadow fescue mixture or for maize in Geipel et al (), and as is shown in our regression analysis, the canopy height is (well) correlated to yield. Slightly lower correlation values ( r = .79) were obtained in comparison with those reported by Viljanen et al () for a timothy and meadow fescue mixture ( r ranges from .79 to .98).…”

“…for barley,Yue et al (2017) for winter wheat,Näsi et al (2018) for a timothy and meadow fescue mixture or for maize inGeipel et al (2014), and as is shown in our regression analysis, the canopy height is (well) correlated to yield. Slightly lower correlation values(r = .79) were obtained in comparison with those reported by Viljanen et al…”

“…This might be due to the problems that could be encountered such as lower canopy or the high variability between the different accessions or cultivars. Compared with previous UAV imagery research in other crops (Geipel et al, ; Raymond et al, ; Li et al, ; Näsi et al, ; Yue et al, ), in this study a higher number of flights (and thus dates), samples and accessions were involved in the estimation of DMY though different regression techniques, from a simple linear regression to more advanced techniques such as partial least squares regression and random forest. We have shown that it is possible to estimate canopy height of perennial ryegrass plots using with a cheap, high‐resolution RGB camera mounted on a drone and predict DMY in a better way (lower RMSE) than with a rising plate meter.…”

“…The mean error calculated in the Z ‐axis was comparable or even better than shown in other studies in grasslands. For example, Näsi et al () or Viljanen et al () presented RMSE values in the Z ‐axis of 6.9 cm and 1.0–4.8 cm respectively. The size of the area “free of vegetation” needed around a GCP depends on the canopy height and the lens used.…”

“…example,Näsi et al (2018) orViljanen et al (2018) presented RMSE values in the Z-axis of 6.9 cm and 1.0-4.8 cm respectively. The size of the area "free of vegetation" needed around a GCP depends on the canopy height and the lens used.…”