摘要
采用近红外光谱技术对辽宁大连、河北遵化、潍坊坊子、潍坊昌邑、潍坊诸城5个产地的100个鸡肉样本进行扫描，对这些样本的近红外光谱进行主成分分析、聚类分析。建立了鸡肉产地溯源的定性判别模型。试验结果表明：在全光谱范围（780～2500nm）内，经二阶求导（13点平滑）和矢量归一化（standard normal variate，SNV）预处理后，5个地区鸡肉的近红外光谱图有显著差异，鸡肉样本的主成分空间分布位于不同的区域，聚类分析树状图中不同产地也各自聚为一类。利用来自5个产区的30个独立样本对模型进行验证，识别率和拒绝率均为100％。此结果表明近红外光谱分析技术可准确、快速追溯鸡肉的产地来源。
One hundred chicken samples from 5 different regions of China were scanned with a near infrared （NIR） spectrophotometer, the chicken samples were subjected to principal component analysis （PCA）, cluster analysis （CA）, and a qualitative model for traceability of chicken was established to discriminant analysis. It was showed that there were significantly differences among the chicken samples from 5 regions based on NIR spectra after the second derivative（13 point smoothing） and standard normal variety（SNV） over the wavelength range of 780nm to 2500 nm. The element contents in the chicken samples were different from different locations, and the different chicken samples can gathered into their own groups in the dendrogram of cluster analysis. This model calibrated by samples was used to pre- dict the varieties of 30 unknown chicken samples, the recognition rate and rejection rate of 100% was achieved. These results demonstrate that NIR fingerprint spectroscopy to trace geographical origin of chicken is accurate and rapid.
出处
《中国食品学报》
EI
CAS
CSCD
北大核心
2014年第12期198-204,共7页
Journal of Chinese Institute Of Food Science and Technology
基金
山东省自主创新专项(2013CXC90202)
关键词
近红外光谱
鸡肉
产地溯源
near infrared spectroscopy
chicken meat
origin traceability
