Title :
The grade recognition of green tea
Author :
Zhang, Rong-xiang ; Wang, Wen-li ; Li, Guang ; Zhao, Xiao-hui ; Zhang, Lian-shui ; Li, Xiao-wei
Author_Institution :
Province Key Lab. for Optoelectron. Inf. Mater. of Heibei, Hebei Univ., Baoding
Abstract :
It is urgent to think up a quick and precise method for the identification of tea varieties and grades. The Fourier-transform infrared spectra of three grades of Xihu Longjing tea are measured, and in the spectra region of 400-4000 cm-1, which are build separately by the Savitzky-Golay smoothing and standard normal variate preprocessing method. It can be shown that the three different grades of Xihu Longjing have nearly the same characteristic peaks in conventional Fourier-transform infrared spectra. And the data analysis shows that the difference of correction coefficient in different grades is small. So only from the spectrum characteristic or correction coefficient the three grades of green tea can not be effectively recognized. The further analysis shows that the average derivation of the spectra can identify effectively and separate the different grades of tea quickly. Based on the analysis, a way of tea recognition which need not relate to the complicated component analysis of tea is given.
Keywords :
Fourier transforms; beverages; infrared spectroscopy; production engineering computing; spectral analysis; Fourier-transform infrared spectra; Savitzky-Golay smoothing; data analysis; green tea grade recognition; Cybernetics; Data analysis; Information science; Infrared detectors; Infrared spectra; Laboratories; Machine learning; Manufacturing; Powders; Spectroscopy; Average deviation; Correlation coefficient; Fourier-transform infrared spectroscopy; Green tea; Recognition;
Conference_Titel :
Machine Learning and Cybernetics, 2008 International Conference on
Conference_Location :
Kunming
Print_ISBN :
978-1-4244-2095-7
Electronic_ISBN :
978-1-4244-2096-4
DOI :
10.1109/ICMLC.2008.4620639