摘要
腊肉中细菌总数(TVC)超标会严重危害人体健康,为寻求一种快速、无损检测腊肉表面细菌总数的方法,利用高光谱成像技术对腊肉的细菌总数进行定量分析。综合比较了多元散射、微分处理等多种预处理方法,最终选定了MSC+标准化进行预处理。并采用区间优化偏最小二乘的方法建立预测模型,得到较好的预测结果,其校正集和预测集的相关系数分别为0.808和0.798,交互验证均方根误差分别为0.115和0.198。实验结果表明:利用高光谱成像技术结合区间iPLS预测模型快速检测腊肉的TVC是可行的。
The total viable count （TVC） in bacon exceeding limits can cause serious damage to human health. In order to find a rapid and nondestructive method of TVC, hyperspectral imaging technique is applied for quantitatively analysis of TVC in bacon. Comprehensively compare the pretreatment method of multiple scattering, derivative methods and so on. Finally, the multiple scattering for pretreatment is used. The interval optimization method of least-squares model is adopted to set up forecast model, and get a good prediction result. The correlation coefficient of calibration and prediction set is 0.808 and 0.798 respectively; the interactive authentication root mean square error is 0. 115 and 0. 198 respectively. Therefore, hyperspectral imaging technique combining with iPLS can be used for the rapid detection of TVC in bacon.
出处
《中国调味品》
CAS
北大核心
2016年第2期74-78,共5页
China Condiment
基金
国家自然科学基金项目(61473009)
北京市自然科学基金项目(4122020)
2015年研究生科研能力提升计划项目
关键词
腊肉
细菌总数
高光谱成像
区间优化最小二乘
bacon
total viable count
hyperspectral imaging
interval optimization least-squares
