Prediction of Albumen pH with Effective Wavelength Regions Selection by Visible-NIR Spectroscopy
The objective of this research was to determine albumen pH of eggs based on visible-near infrared (visible-NIR) diffuse reflectance spectroscopy in the wavelength range of 500-950 nm. One hundred eggs were randomly divided into the calibration and prediction set (60:40). The effective wavelength regions (EWs) were chosen from the spectra pro-processed by first derivative by interval partial least squares (iPLS) and backward interval partial least squares (BiPLS). Then the partial least squares regression (PLSR) and principal component regression (PCR) models were developed with EWs. The performance of the models was evaluated in the prediction set of forty unknown samples. The best models were obtained with the correlation coefficients (Rp) of 0.88 and the root mean square error of prediction (RMSEP) of 0.07 for albumen pH by PLS model in which the EWs were chosen by BiPLS. The results show that the visible-NIR spectroscopy in combination with variable selection was a feasible method to detect albumen pH of egg quality non-destructively.
Visible and Near-Infrared Spectroscopy Albumen pH Egg Quality Backward Interval Partial Least Squares Partial Least Squares Regression
Yande Liu Yanying Peng Xiaoling Dong Xudong Sun Aiguo Ouyang
Institute of Optics-Mechanics-Electronics Technology and Application (OMETA),East China Jiaotong Uni Foreign Languages College, Jiangxi Normal University, Nanchang, 330022, China
国际会议
南昌
英文
1164-1169
2010-10-22（万方平台首次上网日期，不代表论文的发表时间）