First Derivative Transformation Research Articles

Predicting leaf nitrogen content in wolfberry trees by hyperspectral transformation and machine learning for precision agriculture.

Leaf nitrogen content (LNC) is an important indicator for scientific diagnosis of the nutrition status of crops. It plays an important role in the growth, yield and quality of wolfberry. This study aimed to develop new spectral indices (NSIs) and constructed machine learning regression (MLR) models for predicting wolfberry tree LNC. By utilizing four smoothing methods and five mathematic transformation methods, we obtained the original spectral dataset and five spectral transformation datasets for quantitative analysis and model establishment. Subsequently, published vegetation indices (PVIs) were acquired, sensitive wavelengths (SWs) were screened and NSIs were calculated based on SWs. Then MLR models were constructed by combining NSIs from six spectral datasets with three machine learning algorithms. Finally, a comparison was made among the MLR models. The study indicated that the application of mathematical transformation highlighted the differences in spectra, the square root, first-derivative and second-derivative transformation improved the prediction accuracy of MLR models constructed based on NSIs (MLR-NSIs models). However, these transformations had little impact on improving the prediction ability of MLR models constructed based on PVIs (MLR-PVIs models). Additionally, The optimal model for predicting the LNC of wolfberry tree was obtained by using the Random Forest (RF) algorithm combined with NSIs developed by first-derivative transformation spectra. The determination coefficient of validation (Rv2) and ratio of percentage deviation (RPD) was 0.71 and 1.90, respectively. In conclusion, this study has demonstrated that the combination of hyperspectral transformation and machine learning is useful for improving the accuracy of LNC estimation in wolfberry trees.

Quantitative Monitoring of Leaf Area Index in Rice Based on Hyperspectral Feature Bands and Ridge Regression Algorithm

Leaf area index (LAI) is one of the indicators measuring the growth of rice in the field. LAI monitoring plays an important role in ensuring the stable increase of grain yield. In this study, the canopy reflectance spectrum of rice was obtained by ASD at the elongation, booting, heading and post-flowering stages of rice, and the correlations between the original reflectance (OR), first-derivative transformation (FD), reciprocal transformation (1/R), and logarithmic transformation (LOG) with LAI were analyzed. Characteristic bands of spectral data were then selected based on the successive projections algorithm (SPA) and Pearson correlation. Moreover, ridge regression (RR), partial least squares (PLS), and multivariate stepwise regression (MSR) were conducted to establish estimation models based on characteristic bands and vegetation indices. The research results showed that the correlation between canopy spectrum and LAI was significantly improved after FD transformation. Modeling using SPA to select FD characteristic bands performed better than using Pearson correlation. The optimal modeling combination was FD-SPA-VI-RR, with the coefficient of determination (R2) of 0.807 and the root-mean-square error (RMSE) of 0.794 for the training set, R2 of 0.878 and RMSE of 0.773 for the validation set 1, and R2 of 0.705 and RMSE of 1.026 for the validation set 2. The results indicated that the present model may predict the rice LAI accurately, meeting the requirements of large-scale statistical monitoring of rice growth indicators in the field.

Estimation of soil nitrogen in agricultural regions by VNIR reflectance spectroscopy

In the present paper, the novel hyperspectral model was developed for the estimation of Soil Nitrogen (SN) in agricultural lands using Partial Least Squares Regression (PLSR) method. In this regard, an effort has been made on predicting and analyzing SN from several agricultural lands of Phulambri Tehsil of Aurangabad district of Maharashtra, India. The spectra of seventy four (74) agricultural soil samples were acquired between 350–2500 nm by Analytical Spectral Device Field Spec-4 Spectroradiometer under controlled laboratory conditions. The preprocessing was done on acquired spectra by First-derivative Transformation (FDT) and Savitzky–Golay (SG) method for getting suitable information. The PLSR approach was derived from correlation analysis between reflectance spectra and SN features. The resulted coefficient of determination (R2) values was 0.68 and 0.94 before and after pre-treatment with root mean square error of prediction (RMSEP) 4.34 and 1.56, respectively. The identified sensitive wavelength bands of nitrogen content were 480 nm, 511 nm, 653 nm, 997 nm, 1472 nm, 1795 nm, 2210 nm and 2296 nm. In the conclusion, the model is reliable for prediction of SN from agricultural areas. The present research will be useful for decision making in agricultural management.

FT-NIR spectroscopy for non-invasive identification of natural polymers and resins in easel paintings

In the present study, the analytical strengths and limitations of near-infrared (NIR) spectroscopy to non-invasively characterize organic components in painting materials have been investigated. In spite of the increased amount of information available today from advanced modern analytical instrumentations dedicated to cultural heritage, the non-invasive identification of materials belonging to the wide class of organic compounds historically used in paintings is still a challenging task. Near-infrared spectroscopy offers several attractive features that make this technique particularly suitable to this purpose. In fact, it is non-invasive, allows for non-contact measurements in reflectance mode, gives molecular information on complex macromolecules, and can be performed on-site by means of portable devices. First-derivative transformation of reflectance spectroscopic data has been applied to provide a simple and fast way to deduce more information from NIR spectra. This approach has allowed spectral features to be identified that can be useful to distinguish different compounds belonging to the classes of lipids, proteins, and resins. To this purpose, at first, a spectral database of pure standard has been collected. Our analytical approach was then successfully validated on pictorial models reproducing the typical stratigraphy of an easel painting. As final step, the study of a real painting has been attempted and a drying oil, animal glue, and a terpenic natural resin, as well as an earth pigment were clearly identified, as cross-validated by GC-MS analysis.

Measurement of Soluble Solids Contents and pH in Orange Juice Using Chemometrics and Vis−NIRS

The potential of visible and near-infrared reflectance spectroscopy (vis-NIRS) was investigated for its ability to nondestructively detect soluble solids contents (SSC) and pH in orange juices. A total of 104 orange juice samples were used for vis-NIRS at 325-1075 nm using a field spectroradiometer. Wavelet packet transform, standard normal variate transformation (SNV), and Savitzky-Golay first-derivative transformation were applied for the preprocessing of spectral data. The chemometrics of partial least-squares (PLS) regression analysis was performed on the processed spectral data. The evaluation of SSC and pH in orange juices by PLS regression with SNV showed the highest accuracy of the three preprocessing methods. The correlation coefficient (r), standard error of prediction, and the root-mean-square error of prediction for SSC were 0.98, 0.68, and 0.73, respectively, whereas those values for pH were 0.96, 0.06, and 0.06, respectively. The "fingerprint" representing features of orange juices or reflecting sensitivity to some elements at a certain band was proposed on the basis of regression coefficients. It is very useful in the field of food chemistry and further research on other materials. It is concluded that the vis-NIRS technique combined with chemometrics is promising for the fast and nondestructive detection of chemical components in orange juices or other materials.

Length determination of vessel elements in tree trunks used for water and nutrient transport by Fourier transform Raman spectroscopy

Length analysis of vessel elements in tree trunks used for water and nutrient transport is a lengthy, multistep procedure although it reflects environmental stresses on a tree. The feasibility of using FT-Raman spectroscopy for rapid determination of vessel element length in a tree was examined using wood powders of two Eucalyptus species, including samples of various ages and colors. The first-derivative transformation followed by the multiplicative scatter correction of Raman spectroscopic data and the partial least-squares regression revealed highly significant correlation between conventionally measured and Raman-predicted vessel element length with correlation coefficients (r) of 0.843 and 0.826, respectively, in the calibration (for known samples, n = 186) and in the prediction (for unknown samples, n = 40). FT-Raman spectroscopy coupled with multivariate data analysis will contribute to solving the interactions between emerging environmental issues and the anatomical structure of wood, which allow efficient management practices in growing forests to fix atmospheric CO(2) effectively.

Discrimination of Sound and Granary-Weevil-Larva-Infested Wheat Kernels by Near-Infrared Diffuse Reflectance Spectroscopy

Abstract Sound and infested wheat kernels containing lateinstar granary weevil larvae, as identified by X-ray analysis, were used to evaluate the ability of nearinfrared (NIR) spectroscopy to predict the presence of insect larvae in individual wheat kernels. Diffuse reflectance spectra at 1100-2500 nm were recorded from individual infested and sound kernels. Principal component analysis (PCA) of NIR spectra from sound kernels was used to construct calibration models by calculation of Mahalanobis distances. Calibration models were then applied to spectra obtainedfrom both sound and infested kernels in a separate validation set. A 5-factor PCA model using data from a first-derivative spectral transformation was the best model for correctly classifying kernels in an expanded validation sample set, including 100% of sound, 93% of infested, 95% of sound air dried, 86% of infested air dried kernels, and 90% of sound kernels from 6 wheat varieties. Calibrations using the spectral region from 1100 to 1900 nm were least sensitive to kernel moisture differences. Similar results were obtained when discriminant analysis was applied to log 1/R data from selected discrete wavelengths of NIR spectra.

Decomposition of Anthralin in Alkaline Solutions

AbstractUsing conventional absorbance spectra and a first-derivative transformation technique anthralin was shown to undergo apparent first-order decomposition. The pH-log rate profile for the rate of disappearance of anthralin from aqueous solutions was determined at 25°C in the pH range of 7.74 to 10.02 at an ionic strength of 0.5 M. The profile indicated a maximum near the pKa of anthralin. Measured values of the half life of anthralin decreased from 43.9 min at pH 7.74 to 14.8 min at pH 9.44 and then increased to 18.4 min at pH 10.02. No primary kinetic salt effect was observed. Since it was thought likely that trace amounts of metal ions could catalyze the degradation of anthralin, the effect of disodium ethylenediamine-tetraacetic acid addition was also investigated; this chelating agent did not influence the rate. Thermodynamic parameters were calculated from the temperature dependency of the decomposition.

Nonisothermal Aqueous Calorimetry: Computation of Process-Dependent Temperature Change and Aspects of Calorimeter Design

A general method for determining the process-dependent (intrinsic) temperature change in a nonisothermal calorimeter is presented. The nonisothermal approach to calorimetric investigations requires an estimate of the magnitude of the process independent (extrinsic) temperature change during the reaction period. The proposed method can be applied to any calorimeter whose output is a discrete or continuous temperature-time profile. It is based on a first-derivative transformation of the temperature-time profile and the partitioning of the observed temperature variation into two components: pure extrinsic variation, which occurs outside the reaction period, and the combined extrinsic and intrinsic effects during the reaction period. Close examination of the pure extrinsic variation was considered essential, since it provided the basis for identifying the form of a descriptive mathematical function consistent with the observed extrinsic behavior. Once a suitable function was selected, parameters for the equation were determined through a linear regression procedure. The resulting equation was used to predict the extrinsic variation within the reaction period. Subtraction of predicted extrinsic variation from the observed total variation and integration over the time course of the experiment provide an estimate of the process-dependent temperature change. The differential approach was examined for processes performed in a calorimeter of simple design. Aspects of calorimeter design and advantages of the proposed method of data analysis are discussed.