引用本文:
【打印本页】   【下载PDF全文】   查看/发表评论  下载PDF阅读器  关闭
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 2073次   下载 1214 本文二维码信息
码上扫一扫!
分享到: 微信 更多
基于支持向量机的近红外光谱技术鉴别掺假牛肉
张丽华1,2, 相启森1,2, 李顺峰,等3
1.郑州轻工业学院 食品与生物工程学院;2.食品生产与安全河南省协同创新中心;3.河南省农科院 农副产品加工研究所
摘要:
【目的】利用近红外光谱技术对掺入鸭肉的掺假牛肉进行判别研究,为实现掺假牛肉的快速、准确检测提供参考。【方法】市购冷冻储存的牛里脊肉和鸭肉,在牛里脊肉糜中掺入质量分数为0,5%,10%,15%,20%,25%的鸭肉糜(对应m(鸭肉糜)∶m(牛肉糜)分别为0∶20,1∶19,2∶18,3∶17,4∶16和5∶15),制备供试牛肉和掺假牛肉,采用近红外漫反射方式在波数为10 000~4 000 cm-1时采集牛里脊肉、掺假牛里脊肉和鸭肉的近红外光谱,对原始光谱经多元散射校正(Multiplicative scatter correction,MSC)、标准正态变量变换(Standard normal variate transformation,SNV)、面积归一化(Area normalization)、15点平滑处理和一阶导数处理等方法预处理后,对所建支持向量机(nu-SVM)判别模型的预测效果进行比较。【结果】经MSC预处理后所建的牛里脊肉、掺假牛里脊肉和鸭肉的nu SVM判别模型判别性能稳定,对建模集的正确判别率为97.09%,对检验集的正确判别率为94.00%。【结论】近红外光谱技术结合MSC预处理后所建的nu-SVM模型可用于掺杂鸭肉的牛里脊肉的鉴别。
关键词:  掺假牛肉  支持向量机  近红外光谱  鸭肉
DOI:
分类号:
基金项目:郑州轻工业学院博士科研启动基金项目(2013BSJJ003);国家质检总局科技计划项目(2013年度);郑州轻工业学院第三批大学生科技创新项目(2014DC099)
Identification of adulteration of minced beef with near infrared spectroscopy based on support vector machine
ZHANG Lihua,XIANG Qisen,LI Shunfeng,et al
Abstract:
【Objective】A near infrared (NIR) spectroscopy was employed to analyze the adulteration of duck meat in minced beef to realize the nondestructive and quick determination of adulterated beef.【Method】Frozen beef and duck were purchased,and adulterated beef samples were obtained by interfusing different amounts (0,5%,10%,15%,20%,25%) of duck meat (mass ratios of duck to beef were 0∶20,1∶19,2∶18, 3∶17,4∶16 and 5∶15).The original spectra of beef,adulterated beef and duck meat in the wave-number range of 10 000-4 000 cm-1 were obtained by spread reflection NIR.Then,the predicted effects of different support vector machine (SVM) discriminate models that established by original spectra,multiplicative scatter correction (MSC),standard normal variate transformation (SNV),normalization,15 point smooth and 1st derivative were evaluated.【Result】The established model with MSC preprocessing method had the best accuracies of 97.09% and 94.00% for the training set and prediction set,respectively.【Conclusion】The nu-SVM model built by NIR combined with MSC preprocessing method was proved to be a feasible method to identify duck meat in minced beef.
Key words:  adulterated beef  support vector machine  near infrared spectroscopy  duck meat