Using Machine Learning for Estimating Rice Chlorophyll Content from In Situ Hyperspectral Data

Abstract

Chlorophyll is an essential pigment for photosynthesis in crops, and leaf chlorophyll content can be used as an indicator for crop growth status and help guide nitrogen fertilizer applications. Estimating crop chlorophyll content plays an important role in precision agriculture. In this study, a variable, rate of change in reflectance between wavelengths ‘a’ and ‘b’ (RCRWa-b), derived from in situ hyperspectral remote sensing data combined with four advanced machine learning techniques, Gaussian process regression (GPR), random forest regression (RFR), support vector regression (SVR), and gradient boosting regression tree (GBRT), were used to estimate the chlorophyll content (measured by a portable soil–plant analysis development meter) of rice. The performances of the four machine learning models were assessed and compared using root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). The results revealed that four features of RCRWa-b, RCRW551.0–565.6, RCRW739.5–743.5, RCRW684.4–687.1 and RCRW667.9–672.0, were effective in estimating the chlorophyll content of rice, and the RFR model generated the highest prediction accuracy (training set: RMSE = 1.54, MAE =1.23 and R2 = 0.95; validation set: RMSE = 2.64, MAE = 1.99 and R2 = 0.80). The GPR model was found to have the strongest generalization (training set: RMSE = 2.83, MAE = 2.16 and R2 = 0.77; validation set: RMSE = 2.97, MAE = 2.30 and R2 = 0.76). We conclude that RCRWa-b is a useful variable to estimate chlorophyll content of rice, and RFR and GPR are powerful machine learning algorithms for estimating the chlorophyll content of rice.

Corn Biomass Estimation by Integrating Remote Sensing and Long-Term Observation Data Based on Machine Learning Techniques

“…In this study, the performance of the RF model was comparable to that of the XGBoost model, and this finding is in accordance with the results reported by Herrero-Huerta et al [80]. The RF model outperformed the SVC and ANN models, and the results are similar to those of the studies of An et al, Kayah et al, Xu et al, and Zhu et al [14,[81][82][83], which focused on the performance and accuracy of the RF, SVC, ANN and other models, and the RF model was recommended due to its robustness and accuracy in those studies. Previous studies have indicated that the RF model has the ability to resist overfitting and address high-dimensional data [84], which is why the RF model is an effective model in this study.…”

“…The satellite remote sensing technique is a unique and useful method for crop biomass monitoring in a repeatable manner due to the valuable information it provides on vegetation parameters, high spatial coverage and long time series, which effectively compensates for the deficiency of traditional biomass estimation methods [2,8,[11][12][13]. Previous studies have revealed that remote sensing is a reliable and effective technique to obtain biophysical and biochemical crop information [6,[14][15][16][17]. For example, the moderate resolution imaging spectroradiometer (MODIS) of the Earth Observation System (EOS) instrument provides long time series observations at a spatial resolution ranging from 250 to 1000 m in multiple spectral bands at the visible to shortwave infrared (SWIR) wavelengths, with a global coverage of one to two days, which has been widely applied in studies on the variation in vegetation parameters at the regional and global scales [18][19][20].…”

“…Various methods have been developed for crop biomass estimation based on optical remote sensing data, such as statistical, process-based, machine learning, and other methods [21]. Machine learning techniques have become a basic and effective way to model and extract patterns from remote sensing data due to their high computational efficiency, few required variables, and reliable results [14,[21][22][23]. Machine learning techniques, including artificial neural networks (ANNs) [24], random forests (RFs) [25][26][27], support vector machines (SVMs) [28,29], and extreme gradient boosting (XGBoost) [30][31][32] have been widely implemented to evaluate vegetation biomass based on spectral signature parameters [1,17,33,34].…”

Hyperspectral wavelength selection for estimating chlorophyll content of muskmelon leaves

“…However, this strategy might have reduced sample size too much to generate regression models, as RF did not perform as well as in previous studies (Biau & Scornet, 2016). A lot of earlier studies have reported the best machine learning algorithms for estimating leaf chlorophyll contents from hyperspectral reflectance or vegetation indices calculated from reflectance (An et al, 2020;Zhu et al, 2020), however, the combinations of machine learning algorithms and wavelength selection methods were not conducted. The results indicated wavelengths selection is a critical step for chlorophyll content estimation and suitable selection methods made estimation accuracies higher.…”

Assessment of Ultraviolet Impact on Main Pigment Content in Purple Basil (Ocimum basilicum L.) by the Spectrometric Method and Hyperspectral Images Analysis

“…Pigment concentration measurement by the chemical method takes quite a long time, limiting the ability to analyze a large number of samples, so sub-sampling is used instead. Modern digital technologies allow estimating the pigment concentration in a leaf by the analysis of its reflection spectra (analysis of the hyperspectral images) [16][17][18][19]. Based on such data, plant stress states can be determined [20][21][22] as well as the content of substances of interest, for example, shikimic acid [23].…”