基于可见-近红外光谱及随机森林的鸡蛋产地溯源
发布时间：2018-12-12 02:15
【摘要】：为了研究快速无损鉴别鸡蛋产地的可行性,利用可见-近红外光谱技术,采集4种湖北不同产地鸡蛋的透射光谱(500~900 nm),利用中心化、归一化、标准正态变量(SNV)、Savitzky-Golay平滑滤波(SG)和多元散射校正(MSC)、直接正交信号校正(Direct Orthogonal Signal Correction,DOSC)算法对光谱数据进行预处理,采用t分布式随机邻域嵌入(t-distributed stochastic neighbor embedding,t-SNE)、主成分分析(PCA)方法对预处理后的数据降维,并将降维后的数据分别输入极限学习机(extreme learning machine,ELM)和随机森林(random forest,RF),建立鸡蛋产地溯源模型。比较两种方法建立的模型,发现运用DOSC预处理及t-SNE提取的光谱特征信息建立的RF模型鉴别效果最好,训练集和预测集的鉴别正确率分别为100%和98.33%。研究结果表明基于可见-近红外光谱技术对鸡蛋产地溯源是可行的,为进一步研究与开发鸡蛋产地溯源便携式仪器提供技术支持。
[Abstract]:In order to study the feasibility of fast and nondestructive identification of egg origin, the transmission spectra of four kinds of eggs from different areas in Hubei province were collected by using visible near infrared spectroscopy (500 ~ 900 nm),) using centralization, normalization and standard normal variable (SNV),. Savitzky-Golay smoothing filter (SG) and multivariate scattering correction (MSC), direct orthogonal signal correction (Direct Orthogonal Signal Correction,DOSC) algorithm are used to preprocess the spectral data. T distributed random neighborhood embedding (t-distributed stochastic neighbor embedding,t-SNE) is used to preprocess the spectral data. The principal component analysis (PCA) was used to reduce the dimension of pretreated data, and the reduced dimension data were input into the extreme learning machine (extreme learning machine,ELM) and random forest (random forest,RF), respectively, and a traceability model of egg origin was established. Comparing the two models, it is found that the RF model based on DOSC pretreatment and t-SNE extraction is the best, and the accuracy of training set and prediction set are 100% and 98.33%, respectively. The results show that it is feasible to trace the origin of eggs based on visible near infrared spectroscopy, which provides technical support for further research and development of portable instrument for tracing the origin of eggs.
【作者单位】： 华中农业大学工学院;华中农业大学国家蛋品加工技术研发分中心;
【基金】：国家自然科学基金(31371771) 湖北省科技支撑计划项目(2015BBA172) 国家科技支撑计划项目(2015BAD19B05) 公益性行业(农业)科研专项(201303084)
【分类号】：O657.33;TS253.7
，
本文编号：2373703
[Abstract]:In order to study the feasibility of fast and nondestructive identification of egg origin, the transmission spectra of four kinds of eggs from different areas in Hubei province were collected by using visible near infrared spectroscopy (500 ~ 900 nm),) using centralization, normalization and standard normal variable (SNV),. Savitzky-Golay smoothing filter (SG) and multivariate scattering correction (MSC), direct orthogonal signal correction (Direct Orthogonal Signal Correction,DOSC) algorithm are used to preprocess the spectral data. T distributed random neighborhood embedding (t-distributed stochastic neighbor embedding,t-SNE) is used to preprocess the spectral data. The principal component analysis (PCA) was used to reduce the dimension of pretreated data, and the reduced dimension data were input into the extreme learning machine (extreme learning machine,ELM) and random forest (random forest,RF), respectively, and a traceability model of egg origin was established. Comparing the two models, it is found that the RF model based on DOSC pretreatment and t-SNE extraction is the best, and the accuracy of training set and prediction set are 100% and 98.33%, respectively. The results show that it is feasible to trace the origin of eggs based on visible near infrared spectroscopy, which provides technical support for further research and development of portable instrument for tracing the origin of eggs.
【作者单位】： 华中农业大学工学院;华中农业大学国家蛋品加工技术研发分中心;
【基金】：国家自然科学基金(31371771) 湖北省科技支撑计划项目(2015BBA172) 国家科技支撑计划项目(2015BAD19B05) 公益性行业(农业)科研专项(201303084)
【分类号】：O657.33;TS253.7
，
本文编号：2373703
本文链接：https://www.wllwen.com/kejilunwen/huaxue/2373703.html
教材专著
