Application of Wavelet Transform in the Prediction of Chemical Composition of Sunflower Seeds by Near Infrared Reflectance Spectroscopy

Abstract:

Article Preview

In this paper, a method based on wavelet transform, which is used to analyze near infrared spectra, is discussed with the purpose of prediction of the content of oil, crude protein(CP) and moisture in sunflower seeds. By using different decomposing levels of Daubechies 2 wavelet transform, the near infrared spectra signals obtained from 105 intact sunflower seed samples were de-noised. Calibration equations were developed by partial least square regression (PLS) using the reconstructed spectra data with internal cross validation. It was indicated that the prediction effects varied when different wavelet decomposing level were employed. At the wavelet decomposing level 5, the best prediction effect was obtained, with the coefficient of correlation(R)and root mean square error prediction (RMSEP) being 0.953 and 0.466% for moisture;0.963 and 1.259% for crude protein; 0.801 and 1.874% for oil on a dry weight basis. It was concluded that the near infrared spectral model de-noised by means of wavelet transform can be used for the prediction of chemical composition in sunflower seeds for rapid pre-screening of quality characteristics on breeding programs.

Info:

Periodical:

Key Engineering Materials (Volumes 460-461)

Edited by:

Yanwen Wu

Pages:

599-604

DOI:

10.4028/www.scientific.net/KEM.460-461.599

Citation:

R. Z. Han et al., "Application of Wavelet Transform in the Prediction of Chemical Composition of Sunflower Seeds by Near Infrared Reflectance Spectroscopy", Key Engineering Materials, Vols. 460-461, pp. 599-604, 2011

Online since:

January 2011

Export:

Price:

$35.00

In order to see related information, you need to Login.

In order to see related information, you need to Login.