Non-Destructive Identification of the Botanical Origin of Chinese Honey Using Visible/Short Wave-Near Infrared Spectroscopy
The objective of this study was to identify the botanical origins of honeys using visible/short wavenear infrared (Vis/SW-NIR) spectroscopy. Four floral origins, named, acacia (n=109), rape (n=30), longan (n=30) and linden (n=57) were analyzed. In order to acquire an extended-range of honey samples, the acacia honey samples were collected from untreated acacia honeys (n=58), acacia honeys of Zhedafengye (n=25) and acacia honeys of Bainianggongfang (n=26), while the linden honeys were from untreated finden (n=30), of Guanshengyuan (n=11) and of Zhedafengye (n=16). Transmittance mode was adopted as the spectra recording mode and principal component analysis (PCA) together with back propagation artificial neural network (BPANN) was taken as the chemometric methods. After SavitzkyGolay (SG) smoothing of the data, the principal components (PCs) of the PCA extracted from each data set were gathered together into one matrix and analyzed by BPANN. Correct classification of 95％, 100％, 100％ and 100％ were observed for acacia, rape, longan and linden honeys respectively. External prediction was established in order to check the stability of the calibration model, recognition of the new acacia samples was 78.57％. The brand identification test, using the same analysis methods, were lunched to acacia honeys to prove the spectral differences within the botanical origin set of this study and 100％ correct classification was acquired. The obtained results showed that NIR technology combined with chemometric methods such as BPANN could be a suitable and alternative technique to classify the brands and botanical origins of honeys.
Near Infrared Spectroscopy (NIRs) Honey Botanical Origin Brand Identification Principal Component Analysis (PCA) Back Propagation Artificial Neural Network (BPANN) External Prediction Diffuse Reflectance Transflectance Transmittance
Xiangdong Zhao Yong He Yidan Bao
College of Biosystems Engineering and Food Science, Zhejiang University, 268 Kaixuan Road, Hangzhou 310029, P. R. China
国际会议
南昌
英文
1055-1061
2010-10-22（万方平台首次上网日期，不代表论文的发表时间）