These tools will no longer be maintained as of December 31, 2024. Archived website can be found here. PubMed4Hh GitHub repository can be found here. Contact NLM Customer Service if you have questions.


Pubmed for Handhelds

PUBMED FOR HANDHELDS

Search MEDLINE/PubMed

  • Title: [Nondestructive determination of nitrate content in spinach leaves with visible-near infrared high spectra].
    Author: Xue LH, Yang LZ.
    Journal: Guang Pu Xue Yu Guang Pu Fen Xi; 2009 Apr; 29(4):926-30. PubMed ID: 19626874.
    Abstract:
    The objective of the present research was to study the potential of Vis-NIR (visible-near-infrared) high spectra for nondestructive determination of nitrate content in spinach leaves. Five different nitrogen treatments were carried out to achieve a wide range of nitrate content in spinach leaves. The leaf reflectance was measured between 350 to 2,500 nm with a 1 nm step with a leaf clip by ASD Fieldspec FR spectroradiometer, and the nitrate content was measured by spectrophotometric method (National Standard Method of P. R. China). Statistical models were developed using partial least squares (PLS) and principal component regression (PCR) analysis technique, different mathematical treatments for spectra processing such as smoothing, first and second derivative analysis, baseline correction, multiplicative scatter correction (MSC), and standard normal variate correction (SNV), and different wavelength ranges were compared to determine the best model. The dataset was separated into two parts: one used for calibration (n=43), and the other used for test (n=15). First, the model was calibrated and cross-validated with the calibration dataset, then the model was validated with the test dataset to test its prediction ability. The results showed that smoothing, first derivative and second derivative analysis can improve the prediction obviously, while other spectra pre-processing technique e. g. baseline correction, MSC and SNV technique can improve the prediction little. PCR analysis could get better modeling results than PLS analysis. The best model was obtained with the spectra first processed by smoothing then by first derivative change in the full range (350-2,500 nm). Test of the best PLS model and PCR model with an independent dataset exhibited a good agreement between the predicted and observed values, with the correlation coefficient of 0.94 for PLS model and 0.95 for PCR model, and the prediction RMSE was 128.2 mg x kg(-1) for PLS model and 120.8 mg x kg(-1) for PCR model, respectively. These results indicate that visible-NIR spectra technique is a feasible, nondestructive way to predict the nitrate content in spinach leaves.
    [Abstract] [Full Text] [Related] [New Search]