Soil Nutrient Estimation and Mapping in Farmland Based on UAV Imaging Spectrometry

Abstract

Soil nutrient is one of the most important properties for improving farmland quality and product. Imaging spectrometry has the potential for rapid acquisition and real-time monitoring of soil characteristics. This study aims to explore the preprocessing and modeling methods of hyperspectral images obtained from an unmanned aerial vehicle (UAV) platform for estimating the soil organic matter (SOM) and soil total nitrogen (STN) in farmland. The results showed that: (1) Multiplicative Scattering Correction (MSC) performed better in reducing image scattering noise than Standard Normal Variate (SNV) transformation or spectral derivatives, and it yielded a result with higher correlation and lower signal-to-noise ratio; (2) The proposed feature selection method combining Successive Projections Algorithm (SPA) and Competitive Adaptive Reweighted Sampling algorithm (CARS), could provide selective preference for hyperspectral bands. Exploiting this method, 24 and 22 feature bands were selected for SOM and STN estimation, respectively; (3) The particle swarm optimization (PSO) algorithm was employed to obtain optimized input weights and bias values of the extreme learning machine (ELM) model for more accurate prediction of SOM and STN. The improved PSO-ELM model based on the selected preference bands achieved higher prediction accuracy (R2 of 0.73 and RPD of 1.91 for SOM, R2 of 0.63, and RPD of 1.53 for STN) than support vector machine (SVM), partial least squares regression (PLSR), and the ELM model. This study provides an important guideline for monitoring soil nutrient for precision agriculture with imaging spectrometry.

Predicting Soil Organic Matter, Available Nitrogen, Available Phosphorus and Available Potassium in a Black Soil Using a Nearby Hyperspectral Sensor System

“…The model based on samples of 30% SM, which is close to real soil moisture levels after irrigation in the field, showed a robust performance with the smallest RMSEP fluctuation and achieved a suitable performance with three components. Due to the presence of functional groups such as C-H, -COOH, -OH and N-H in soil organic compounds corresponding to spectral response at different wavelengths, spectral characteristics could be explained by different component numbers [9]. Thus, a model with more components is expected to describe more SOM features, improving the accuracy.…”

“…Hyperspectral imaging (HSI) is a non-destructive method that can provide detailed and highly resolved reflectance characteristics of target materials on different scales, and it has the advantage of capturing both spatial and spectral information [8]. Currently, spectrometry in the laboratory has been widely applied to the quantitative inversion of SOM based on the organic matter-sensitive bands that exist in the visible range of 550-770 nm and the near-infrared range of 1300-1500 nm [4,9]. Yu et al [10] found that the correlation coefficients of soil organic carbon with the bands in the near-infrared wavelengths of 747 to 1000 nm and 1010 to 1136 nm were significant at the significance level of 0.01.…”

“…For similar near-infrared wavelengths, the correlations of total phosphorus, total nitrogen and total potassium were significant at the significance level of 0.05. Yang, Bao, Li, Liu, Fu and Mao [9] found that unmanned aerial vehicle (UAV) HSI can be exploited to quantify SOM and soil total nitrogen in farmland. This technique was therefore successfully applied to evaluate the spatial variability of SOM and other soil properties within both small profiles and on a large scale.…”

A Novel Brain Mapping Method without Predetermined Targets Using Hyperspectral Microscopy

“…The spectrum recorded by hyperspectral imaging has high wavelength resolution and covers a wide range of wavelengths. Hyperspectral cameras have been used in chemistry and biology [38][39][40][41][42] , for example in the measurement of food freshness 43,44 and in the automatic monitoring of farm products [45][46][47] . Hyperspectral imaging has also been applied in neuroscience, such as for real-time metabolic and hemodynamic monitoring of the brain in vivo [48][49][50][51] .…”

Evaluation of Different Methods and Models for Grass Cereals’ Production Estimation: Case Study in Wheat

“…Soil and plant analysis are basic techniques (classical methods) for assessing soil fertility, nutrient available content for plant nutrition, and fertilizer interventions [22,23], but they are disadvantageous in terms of time, human resources, and costs. At the same time, imaging methods based on remote sensing, UAV, or terrestrial images provide very accurate real-time information on soil quality and fertility [24,25], crops' response to fertilizer application [26,27], and plants' nutrition status [10,28,29]. The use of vegetation indices by imaging analysis can remove or mitigate disturbing factors in crop evaluation, such as differences in plant species, canopy coverage, soil background, atmospheric condition, illumination, shadowing, solar angle, etc.…”