True Accuracy of near Infrared Spectroscopy and its Dependence on Precision of Reference Data

2002 ◽  
Vol 10 (1) ◽  
pp. 15-25 ◽  
Author(s):  
L.K. Sørensen
Keyword(s):  
Near Infrared ◽  
Cross Validation ◽  
Reference Data ◽  
Nir Spectroscopy ◽  
Limited Resources ◽  
Standard Errors ◽  
Calibration Data ◽  
Pls Regression ◽  
Calibration Equations ◽  
Very High

A more precise estimate of the accuracy of near infrared (NIR) spectroscopy is obtained when the measured standard errors of cross validation ( SECV) and prediction ( SEP) are corrected for imprecision of the reference data. The significance of correction increases with increasing imprecision of reference data. Very high precision of reference data obtained through replicate analyses under reproducibility conditions may not be the optimal goal for the development of calibration equations. In a situation of limited resources, the precision of the reference data should be related to the obtainable accuracy of the spectroscopic system. Investigation of several routine applications based on the partial least-squares (PLS) regression technique showed that increased precision of calibration data only resulted in marginal improvements in true accuracy if the total standard error of reference results from the beginning was less than the estimated true accuracy of the corresponding NIR calibration.

Fourier Transform near Infrared Spectroscopy Applied to Analysis of Chocolate

2000 ◽  
Vol 8 (4) ◽  
pp. 251-257 ◽  
Author(s):  
Jolana Tarkošová ◽  
Jana Čopíková
Keyword(s):  
Fourier Transform ◽  
Reference Values ◽  
Near Infrared ◽  
Cross Validation ◽  
Nir Spectroscopy ◽  
Sufficient Accuracy ◽  
Standard Methods ◽  
Calibration Models ◽  
Predicted Values ◽  
Calibration Equations

Fourier transform near infrared (FT-NIR) spectroscopy was used to establish calibration equations with the aim of determining sucrose, lactose, fat and moisture in chocolate. The possibility of using FT-NIR spectroscopy for evaluating rheological properties (viscosity and yield) of chocolate was also investigated. The concentrations of individual components and the values of viscosity and yield obtained by standard methods were used as reference values. The spectra of 96 chocolate samples were recorded in reflectance mode in the range of 910–2500 nm using an FT-NIR Nicolet Avatar 360N spectrometer equipped with the UpDRIFT accessory. The first or second derivative transformation of the original NIR spectra gave the best accuracy. A partial least squares (PLS) algorithm was used to create calibration models relating reference values to spectral data. The models were validated using cross-validation. The validation results proved that fat, sucrose and lactose can be predicted with sufficient accuracy, while predicted values for moisture, viscosity and yield are less reliable.

scholarly journals Feasibility of Using VIS/NIR Spectroscopy and Multivariate Analysis for Pesticide Residue Detection in Tomatoes

Processes ◽  
2021 ◽  
Vol 9 (2) ◽  
pp. 196
Author(s):  
Araz Soltani Nazarloo ◽  
Vali Rasooli Sharabiani ◽  
Yousef Abbaspour Gilandeh ◽  
Ebrahim Taghinezhad ◽  
Mariusz Szymanek ◽  
...  
Keyword(s):  
Near Infrared ◽  
Cross Validation ◽  
Low Cost ◽  
Moving Average ◽  
Nir Spectroscopy ◽  
Control Treatment ◽  
Residue Detection ◽  
Non Destructive ◽  
Reflective Spectroscopy

The purpose of this work was to investigate the detection of the pesticide residual (profenofos) in tomatoes by using visible/near-infrared spectroscopy. Therefore, the experiments were performed on 180 tomato samples with different percentages of profenofos pesticide (higher and lower values than the maximum residual limit (MRL)) as compared to the control (no pesticide). VIS/near infrared (NIR) spectral data from pesticide solution and non-pesticide tomato samples (used as control treatment) impregnated with different concentrations of pesticide in the range of 400 to 1050 nm were recorded by a spectrometer. For classification of tomatoes with pesticide content at lower and higher levels of MRL as healthy and unhealthy samples, we used different spectral pre-processing methods with partial least squares discriminant analysis (PLS-DA) models. The Smoothing Moving Average pre-processing method with the standard error of cross validation (SECV) = 4.2767 was selected as the best model for this study. In addition, in the calibration and prediction sets, the percentages of total correctly classified samples were 90 and 91.66%, respectively. Therefore, it can be concluded that reflective spectroscopy (VIS/NIR) can be used as a non-destructive, low-cost, and rapid technique to control the health of tomatoes impregnated with profenofos pesticide.

Rapid Determination of Heartwood Extractives in Larix sp. by Means of Fourier Transform near Infrared Spectroscopy

2002 ◽  
Vol 10 (3) ◽  
pp. 203-214 ◽  
Author(s):  
N. Gierlinger ◽  
M. Schwanninger ◽  
B. Hinterstoisser ◽  
R. Wimmer
Keyword(s):  
Fourier Transform ◽  
Near Infrared ◽  
Cross Validation ◽  
Nir Spectroscopy ◽  
Solid Wood ◽  
Hot Water ◽  
Extractive Content ◽  
Fast Method ◽  
Wood Powder ◽  
Test Set

The feasibility of Fourier transform near infrared (FT-NIR) spectroscopy to rapidly determine extractive and phenolic content in heartwood of larch trees ( Larix decidua MILL., L. leptolepis (LAMB.) CARR. and the hybrid L. x eurolepis) was investigated. FT-NIR spectra were collected from wood powder and solid wood using a fibre-optic probe. Partial Least Squares (PLS) regression analyses were carried out describing relationships between the data sets of wet laboratory chemical data and the FT-NIR spectra. Besides cross and test set validation the established models were subjected to a further evaluation step by means of additional wood samples with unknown extractive content. Extractive and phenol contents of these additional samples were predicted and outliers detected through Mahalanobis distance calculations. Models based on the whole spectral range and without data pre-processing performed well in cross-validation and test set validation, but failed in the evaluation test, which is based on spectral outlier detection. But selection of data pre-processing methods and manual as well as automatic restriction of wavenumber ranges considerably improved the model predictability. High coefficients of determination ( R2) and low root mean square errors of cross-validation ( RMSECV) were obtained for hot water extractives ( R2 = 0.96, RMSECV = 0.86%, range = 4.9–20.4%), acetone extractives ( R2 = 0.86, RMSECV = 0.32%, range = 0.8–3.6%) and phenolic substances ( R2 = 0.98, RMSECV = 0.21%, range = 0.7–4.9%) from wood powder. The models derived from wood powder spectra were more precise than those obtained from solid wood strips. Overall, NIR spectroscopy has proven to be an easy to facilitate, reliable, accurate and fast method for non-destructive wood extractive determination.

Ability of Vis-PIR spectroscopy to monitor changes in organic carbon of loamy soils at two depths

2021 ◽  
Author(s):  
Hayfa Zayani ◽  
Youssef Fouad ◽  
Didier Michot ◽  
Zeineb Kassouk ◽  
Zohra Lili-Chabaane ◽  
...  
Keyword(s):  
Organic Carbon ◽  
Near Infrared ◽  
Full Range ◽  
Nir Spectroscopy ◽  
Soil Samples ◽  
Calibration Data ◽  
Least Squares Regression ◽  
Data Set ◽  
Soil Carbon Stocks ◽  
The Mean

<p>Visible-Near Infrared (Vis-NIR) spectroscopy has proven its efficiency in predicting several soil properties such as soil organic carbon (SOC) content. In this preliminary study, we explored the ability of Vis-NIR to assess the temporal evolution of SOC content. Soil samples were collected in a watershed (ORE AgrHys), located in Brittany (Western France). Two sampling campaigns were carried out 5 years apart: in 2013, 198 soil samples were collected respectively at two depths (0-15 and 15-25 cm) over an area of 1200 ha including different land use and land cover; in 2018, 111 sampling points out of 198 of 2013 were selected and soil samples were collected from the same two depths. Whole samples were analyzed for their SOC content and were scanned for their reflectance spectrum. Spectral information was acquired from samples sieved at 2 mm fraction and oven dried at 40°C, 24h prior to spectra acquisition, with a full range Vis-NIR spectroradiometer ASD Fieldspec®3. Data set of 2013 was used to calibrate the SOC content prediction model by the mean of Partial Least Squares Regression (PLSR). Data set of 2018 was therefore used as test set. Our results showed that the variation ∆SOC<sub>obs</sub><sub></sub>obtained from observed values in 2013 and 2018 (∆SOC<sub>obs</sub> = Observed SOC (2018) - Observed SOC (2013)) is ranging from 0.1 to 25.9 g/kg. Moreover, our results showed that the prediction performance of the calibrated model was improved by including 11 spectra of 2018 in the 2013 calibration data set (R²= 0.87, RMSE = 5.1 g/kg and RPD = 1.92). Furthermore, the comparison of predicted and observed ∆SOC between 2018 and 2013 showed that 69% of the variations were of the same sign, either positive or negative. For the remaining 31%, the variations were of opposite signs but concerned mainly samples for which ∆SOCobs is less than 1,5 g/kg. These results reveal that Vis-NIR spectroscopy was potentially appropriate to detect variations of SOC content and are encouraging to further explore Vis-NIR spectroscopy to detect changes in soil carbon stocks.</p>

scholarly journals Application of FT near spectroscopy for determination of true protein and casein in milk

2008 ◽  
Vol 52 (No. 9) ◽  
pp. 284-291 ◽  
Author(s):  
K. Šustová ◽  
J. Růžičková ◽  
J. Kuchtík
Keyword(s):  
Cross Validation ◽  
Reference Data ◽  
Correlation Coefficients ◽  
Nir Spectroscopy ◽  
Rapid Analysis ◽  
Validation Method ◽  
Goat’S Milk ◽  
Goat's Milk ◽  
True Protein

Our study deals with a possibility of determining true protein and casein in cow’s, ewe’s and goat’s milk and in ewe’s colostrums by FT NIR spectroscopy. Samples of milk were analysed by FT NIR in the reflectance mode with the transflectance cuvette. The values of correlation coefficients of calibration were as follows: cow’s protein 0.943; cow’s casein 0.964; ewe’s protein 0.997; ewe’s casein 0.977; goat’s protein 0.989; goat’s casein 0.890; ewe’s colostrum protein 0.983. Calibration was tested using the same set of samples by the cross validation method. The values of correlation coefficients of validation were as follows: cow’s protein 0.923; cow’s casein 0.910; ewe’s protein 0.994; ewe’s casein 0.963; goat’s protein 0.972; goat’s casein 0.814; ewe’s colostrum protein 0.871. The NIRS results were compared with reference data and no significant differences between them were found (<i>P</i> = 0.05). Results of this study indicate that FT NIR spectroscopy can be used for a rapid analysis of protein and casein in cow’s, ewe’s and goat’s milk and ewe’s colostrum.

Lipid Analysis by near Infrared Spectroscopy to Evaluate Inhibitory Effects of Swine Waste Compost on Plant Growth

2008 ◽  
Vol 16 (5) ◽  
pp. 481-486 ◽  
Author(s):  
Takayuki Fujiwara ◽  
Keiichi Murakami
Keyword(s):  
Plant Growth ◽  
Lipid Content ◽  
Near Infrared ◽  
Nir Spectroscopy ◽  
Swine Waste ◽  
Inhibitory Effects ◽  
Absorption Bands ◽  
Freeze Dried ◽  
Calibration Equations

The lipid content of swine manure decreases during the process of composting, and inhibitory effects of compost on root growth in germination tests are strongly correlated to lipid content. Therefore, we tested whether the determination of the lipid content of swine waste compost by near infrared (NIR) spectroscopy provided a measure by which the degree of inhibition of plant growth by immature compost could be predicted. Reflectance spectra of untreated compost samples, as well as freeze-dried and milled samples, were taken using a scanning monochromator. Second derivative spectra from 700 nm to 2500 nm and multiple regression analysis were used to develop calibration equations for lipid content and moisture. A pronounced absorption peak of lipid was found at 2310 nm, attributable to the absorption bands of the CH2 stretching–bending combination. However, calibration equations containing this absorption band were inappropriate for lipid determination, because sawdust and rice husk, which were added to the compost, influenced the spectra in this band. The standard error of prediction ( SEP) of the best calibrations for lipids in dry and untreated samples was 6.0 g kg−1 and 3.2 g kg−1, while the ratios of the standard deviation and the range in the prediction set to SEP (RPD and RER) were 5.5 and 2.8, and 13.5 and 5.0, respectively. The main wavelengths of these calibration equations were 1700 nm for dry samples and 1764 nm for untreated samples, which were attributed to the absorption bands of the CH2 stretching first overtone. In conclusion, the determination of lipid content in dry compost samples by NIR spectroscopy provided an indirect estimate of the maturity of swine waste compost. Moreover, NIR spectroscopy was found useful for the rough assessment of the maturity of untreated swine waste compost.

scholarly journals Characteristics of fermented dough predicted by using the NIR technique

2011 ◽  
Vol 23 (No. 5) ◽  
pp. 184-189 ◽  
Author(s):  
O. Jirsa ◽  
M. Hrušková
Keyword(s):  
Correlation Coefficient ◽  
Cross Validation ◽  
Rheological Characteristics ◽  
Pls Regression ◽  
Independent Validation ◽  
Predicted Values ◽  
Accuracy Of Prediction ◽  
Gas Volume ◽  
Calibration Equations

Rheological characteristics of wheat doughs prepared from 179 variety and commercial wheat flour samples (crop years 1999, 2001, 2002, 2003 and 2004) were assessed with SJA fermentograph (Sweden). NIR spectra of the samples were measured on the spectrometer NIRSystem 6500. Calibration equations with cross and independent validations for all rheological characteristics were computed by NIR Software ISI Present WINISI II using mPLS/PLS regression and ANN. The quality of prediction was evaluated by means of coefficients of correlation between the measured and the predicted values from cross and independent validation. A statistically significant dependence (with probability higher than 99%) was determined in all mentioned rheological characteristics in the case of cross-validation. The results suggested the possibility of prediction particularly for the fermentation gas volume parameter, where a high correlation was achieved (r = 0.888). The accuracy of prediction by independent validation was not found on such a satisfactory level. For the sets of ten commercial flours, higher correlation coefficient was calculated for the final dough volume parameter. &nbsp;

The Use of near Infrared Spectroscopy to Quantify Lignite-Derived Carbon in Humus-Lignite Mixtures

2007 ◽  
Vol 15 (3) ◽  
pp. 195-200 ◽  
Author(s):  
Marcin Chodak ◽  
Maria Niklińska ◽  
Friedrich Beese
Keyword(s):  
Near Infrared ◽  
Regression Coefficient ◽  
Nir Spectroscopy ◽  
Principal Component ◽  
Mine Soils ◽  
Organic Horizons ◽  
Predicted Values ◽  
Calibration Equations ◽  
Mixed Samples ◽  
Forest Humus

Assessment of the percentage of lignite-derived C (lign-C%) in mine soils may be achieved only by using time-consuming and expensive methods. The objectives of this study were (1) to compare near infrared (NIR) spectra of forest humus and lignite and (2) to test whether NIR spectroscopy may assess lign-C% in artificial mixtures of humus and lignite. The experiment consisted of three trials (T1, T2 and T3). In T1 the mixed samples ( n = 75) were produced from one humus sample and one lignite sample, in T2 (n = 74) from 74 different humus samples and one lignite sample and in T3 (n = 74) from 74 different humus samples and 15 lignite samples. In each trial, 35 samples were used to develop calibration equations and the remaining samples were used for validation. The humus and the lignite samples used to produce the mixed samples were analysed for C, H, N and S and their NIR spectra were recorded. The lignite samples contained more C, H and S and less N than the humus samples. Principal component analysis revealed significant differences between NIR spectra of the humus and the lignite samples. The prediction of lign-C% in T1 [regression coefficient (b) of linear regression (measured against predicted values) = 0.99, correlation coefficient ( r2) = 1.00, standard error of prediction (SEP) = 1.2%] and T2 ( b = 0.99, r2 = 0.99, SEP = 1.9%) was very good and in T3 satisfactory ( b = 0.83, r2 = 0.92, SEP = 4.0% ). The calibration equations of T2 predicted lign-C% satisfactorily and also in the validation samples of T3 (b = 0.88, r2 = 0.93, SEP = 4.0% ). The results indicate the ability of NIR spectroscopy to predict lign-C% in the mixed humus and lignite samples and suggest usefulness of NIR spectroscopy for the assessment of the percentage of lignite-derived C in the organic horizons of mine soils.

Estimating the Composition of Tomato Juice Products by near Infrared Spectroscopy

2003 ◽  
Vol 11 (2) ◽  
pp. 123-136 ◽  
Author(s):  
Athanasia M. Goula ◽  
Konstantinos G. Adamopoulos
Keyword(s):  
Near Infrared ◽  
Correlation Coefficients ◽  
Nir Spectroscopy ◽  
Tomato Juice ◽  
Calibration Data ◽  
Multilinear Regression ◽  
Statistical Parameters ◽  
Data Set ◽  
Prediction Ability ◽  
Specific Product

The use of near infrared (NIR) reflectance spectroscopy for the rapid and accurate measurement of moisture, sugar, acid, protein and salt was explored in a diverse group of tomato juice products. Partial and overall calibrations were performed on four different tomato juice products. Partial calibrations for each product included samples of the specific product, whereas overall calibration used samples of all the products. Samples were analysed employing traditional chemical methods and scanned using an Instalab 600-Dickey-John NIR apparatus to obtain NIR spectra. Calibrations were achieved with the use of multilinear regression between chemical and spectral data from each calibration data set. A separate set of samples was used to validate the calibrations. Linear regression was applied to compare the results obtained by NIR spectroscopy for all constituents of the validation set with those obtained by the reference methods. In addition, the root mean square error of prediction ( RMSEP), the bias and the correlation coefficients ( r and r′) were calculated. All of the statistical parameters were better with overall than with partial calibrations. Prediction ability of overall calibration was very good for all the constituents. r and r′ values were higher than 0.9488 and 0.9453, respectively, RMSEP values were smaller than 0.1067, whereas bias varied from −0.020 to 0.016. The partial calibrations are considerable less variable with the correlation coefficients r and r′ ranged from 0.8890 to 0.9477 and from 0.7202 to 0.8518, respectively, RMSEP varied from 0.0647 to 0.4942 and bias from −0.365 to 0.071. NIR measurement as performed by the Dickey-John Analyser was proved a rapid and accurate method for analysis of tomato juice samples and may be used as a replacement for conventional expensive and time-consuming wet chemistry methods.

Sign in / Sign up

Export Citation Format

Share Document