摘要
苹果的品种对于苹果汁的质量是极其重要的,因此检测苹果汁的品种产地对于打击假冒伪劣产品,维护消费者的合法权益是十分有必要的。本文采集了云南和陕西两个区域的不同品种的苹果所榨的89份苹果汁的荧光光谱,其中62个样品作为建模集利用支持向量机进行了建模并进行交叉验证,识别准确率为100%。剩余的27个样品作为验证集对模型进行验证,识别准确率为96.3%。实验结果证明荧光光谱结合支持向量机技术是一种可靠的检测果汁品种产地的技术手段。
The variety of the raw fruit is important for the quality of fruit juice.To punish the selling of fake apple juice and protect the legal rights of the consumers,it is necessary to discriminate the variety of apple juice.In this work,totally,89 fluorescent spectra of the apple juice samples,which were produced by two kinds of apples from Xi’an and Yunnan,were measured.Among these fluorescent spectra,62 fluorescent spectra were used to build the model based on support vector machine and the model was cross validation with leave one out cross validation method.The accuracy of the cross validation was 100%.The rest 27 fluorescent spectra were used to validate the model and the accuracy was 96.3%.All these results demonstrated that fluorescent spectroscopy combined with support vector machine is feasible to discriminate the variety of the apple juice.
作者
盛立哲
季仁东
王晓燕
姚华
于银山
魏丹丹
卞海溢
SUN Nan
SHENG Lizhe;JI Rendong;WANG Xiaoyan;YAO Hua;YU Yinshan;WEI Dandan;SUN Nan;BIAN Haiyi(Faculty of Electronic Information Engineering,Huaiyin Institute of Technology,Huai’an,Jiangsu,223003,China;Jiangsu Engineering Laboratory for Lake Environment Remote Sensing Technologies,Huai’an,Jiangsu,223003,China)
出处
《光散射学报》
2022年第3期244-248,共5页
The Journal of Light Scattering
基金
国家自然科学基金(61801188,62141502)
淮安市自然科学研究计划(HAB201902)。
关键词
荧光光谱
支持向量机
苹果汁
分类
fluorescent spectrum
support vector machine
apple juice
discrimination