Rapid estimation of soil water content based on hyperspectral reflectance combined with continuous wavelet transform, feature extraction, and extreme learning machine.
Soil water content is one of the critical indicators in agricultural systems. Visible/near-infrared hyperspectral remote sensing is an effective method for soil water estimation. However, noise removal from massive spectral datasets and effective feature extraction are challenges for achieving accurate soil water estimation using this technology.
This study proposes a method for hyperspectral remote sensing soil water content estimation based on a combination of continuous wavelet transform (CWT) and competitive adaptive reweighted sampling (CARS). Hyperspectral data were collected from soil samples with different water contents prepared in the laboratory. CWT, with two wavelet basis functions (mexh and gaus2), was used to pre-process the hyperspectral reflectance to eliminate noise interference. The correlation analysis was conducted between soil water content and wavelet coefficients at ten scales. The feature variables were extracted from these wavelet coefficients using the CARS method and used as input variables to build linear and non-linear models, specifically partial least squares (PLSR) and extreme learning machine (ELM), to estimate soil water content.
The results showed that the correlation between wavelet coefficients and soil water content decreased as the decomposition scale increased. The corresponding bands of the extracted wavelet coefficients were mainly distributed in the near-infrared region. The non-linear model (ELM) was superior to the linear method (PLSR). ELM demonstrated satisfactory accuracy based on the feature wavelet coefficients of CWT with the mexh wavelet basis function at a decomposition scale of 1 (CWT(mexh_1)), with R2, RMSE, and RPD values of 0.946, 1.408%, and 3.759 in the validation dataset, respectively. Overall, the CWT(mexh_1)-CARS-ELM systematic modeling method was feasible and reliable for estimating the water content of sandy clay loam.
Chen S
，Gao J
，Lou F
，Tuo Y
，Tan S
，Shan Y
，Luo L
，Xu Z
，Zhang Z
，Huang X
... -
《PeerJ》
Improved multivariate modeling for soil organic matter content estimation using hyperspectral indexes and characteristic bands.
Soil organic matter (SOM) is a key index of soil fertility. Calculating spectral index and screening characteristic band reduce redundancy information of hyperspectral data, and improve the accuracy of SOM prediction. This study aimed to compare the improvement of model accuracy by spectral index and characteristic band. This study collected 178 samples of topsoil (0-20 cm) in the central plain of Jiangsu, East China. Firstly, visible and near-infrared (VNIR, 350-2500 nm) reflectance spectra were measured using ASD FieldSpec 4 Std-Res spectral radiometer in the laboratory, and inverse-log reflectance (LR), continuum removal (CR), first-order derivative reflectance (FDR) were applied to transform the original reflectance (R). Secondly, optimal spectral indexes (including deviation of arch, difference index, ratio index, and normalized difference index) were calculated from each type of VNIR spectra. Characteristic bands were selected from each type of spectra by the competitive adaptive reweighted sampling (CARS) algorithm, respectively. Thirdly, SOM prediction models were established based on random forest (RF), support vector regression (SVR), deep neural networks (DNN) and partial least squares regression (PLSR) methods using optimal spectral indexes, denoted here as SI-based models. Meanwhile, SOM prediction models were established using characteristic wavelengths, denoted here as CARS-based models. Finally, this research compared and assessed accuracy of SI-based models and CARS-based models, and selected optimal model. Results showed: (1) The correlation between optimal spectral indexes and SOM was enhanced, with absolute value of correlation coefficient between 0.66 and 0.83. The SI-based models predicted SOM content accurately, with the coefficient of determination (R2) and root mean square error (RMSE) values ranging from 0.80 to 0.87, 2.40 g/kg to 2.88 g/kg in validation sets, and relative percent deviation (RPD) value between 2.14 and 2.52. (2) The accuracy of CARS-based models differed with models and spectral transformations. For all spectral transformations, PLSR and SVR combined with CARS displayed the best prediction (R2 and RMSE values ranged from 0.87 to 0.92, 1.91 g/kg to 2.56 g/kg in validation sets, and RPD value ranged from 2.41 to 3.23). For FDR and CR spectra, DNN and RF models achieved more accuracy (R2 and RMSE values ranged from 0.69 to 0.91, 1.90 g/kg to 3.57 g/kg in validation sets, and RPD value ranged from 1.73 to 3.25) than LR and R spectra (R2 and RMSE values from 0.20 to 0.35, 5.08 g/kg to 6.44 g/kg in validation sets, and RPD value ranged from 0.96 to 1.21). (3) Overall, the accuracy of SI-based models was slightly lower than that of CARS-based models. But spectral index had a good adaptability to the models, and each SI-based model displayed the similar accuracy. For different spectra, the accuracy of CARS-based model differed from modeling methods. (4) The optimal CARS-based model was model CARS-CR-SVR (R2 and RMSE: 0.92 and 1.91 g/kg in validation set, RPD: 3.23). The optimal SI-based model was model SI3-SVR (R2 and RMSE: 0.87 and 2.40 g/kg in validation set, RPD: 2.57) and model SI-SVR (R2 and RMSE: 0.84 and 2.63 g/kg in validation set, RPD: 2.35).
Zhao MS
，Wang T
，Lu Y
，Wang S
，Wu Y
... -
《PLoS One》
Hyperspectral inversion of heavy metal content in reclaimed soil from a mining wasteland based on different spectral transformation and modeling methods.
Conventional methods for investigating heavy metal contamination in soil are time consuming and expensive. We explored reflectance spectroscopy as an alternative method for assessing heavy metals. Four spectral transformation methods, first-order differential (FDR), second-order differential (SDR), continuum removal (CR) and continuous wavelet transform (CWT), are used for the original spectral data. Spectral preprocessing effectively eliminated the noise and baseline drifting and also highlighted the locations of the spectral feature bands. Partial least squares regression (PLSR) and radial basis function neural network (RBF) were used to study the hyperspectral inversion of four heavy metals (Cr, As, Ni, Cd). The inversion models of four heavy metals were established in the bands with the highest correlation coefficient. The inversion effects were evaluated by the coefficient of determination (R2), root mean square error (RMSE) and residual predictive deviation (RPD) indexes. The R values of the correlation coefficient were significantly improved after smoothing and spectral transformation compared to the original waveband. The method combining continuous wavelet transform (CWT) with radial basis function neural network (RBF) had the best inversion effect on the four heavy metals. When compared to partial least squares regression (PLSR), the RMSE values were reduced by approximately 2. The CWT-RBF method can be used as a means of inversion of heavy metals in mining wasteland reclaimed land.
Zhang S
，Shen Q
，Nie C
，Huang Y
，Wang J
，Hu Q
，Ding X
，Zhou Y
，Chen Y
... -
《-》
ResearchGate
(全网免费下载)
钛学术
(全网免费下载)
