基于近红外高光谱成像技术的干制红枣品种鉴别

发布时间：2018-02-16 19:01

本文关键词： 近红外高光谱成像干制红枣鉴别纹理特征特征融合　出处：《光谱学与光谱分析》2017年03期 　论文类型：期刊论文

【摘要】：为实现干制红枣的快速鉴别,提出了一种基于近红外高光谱成像技术的鉴别方法。采集四个品种共240个样本干制红枣的近红外高光谱图像(1 000~1 600 nm)。通过主成分分析法(principal component analysis,PCA)、载荷系数法(x-Loading Weights,x-LW)和连续投影算法(successive projections algorithm,SPA)分别提取7个、8个和10个特征波长;基于灰度共生矩阵(gray level co-occurrence matrix,GLCM)提取第一主成分图像的纹理特征。分别以光谱特征、纹理特征、光谱和纹理融合特征作为输入,建立偏最小二乘判别分析(partial least squares-discriminant analysis,PLS-DA)、反向传播神经网络(back-propagation neural network,BPNN)和最小二乘支持向量机(least squares support vector machines,LS-SVM)模型。结果显示,基于融合特征的模型鉴别率高于分别基于光谱特征或纹理特征的模型鉴别率;基于融合特征的BPNN模型的结果最优,对预测集样本鉴别正确率为100%。说明近红外高光谱成像技术可用于干制红枣品种的快速鉴别。
[Abstract]:In order to realize the rapid identification of dried jujube, A method of identification based on near infrared hyperspectral imaging (NIR) was proposed. Four varieties of dried jujube were collected from 240 samples of red jujube. The method of principal component analysis (PCA) and loading coefficient method (PCAA) and the method of loading coefficient were used to collect 1 000 ~ 1 000 NIR hyperspectral images of Chinese jujube (Zizyphus jujuba L.) by principal component analysis (PCA). Successive projections algorithm (spa) extracts 7, 8 and 10 characteristic wavelengths, respectively. The texture feature of the first principal component image is extracted based on gray level co-occurrence matrix GLCM. The spectral feature, texture feature, spectral feature and texture fusion feature are used as input, respectively. The partial least squares-discriminant analysis (PLS-DAN), back-propagation neural network (BPNN) and least squares support vector machine least squares support vector machine (LS-SVM) models are established. The discriminant rate of model based on fusion feature is higher than that based on spectral feature or texture feature, and the result of BPNN model based on fusion feature is optimal. The accuracy rate of sample identification for prediction set is 100. It is suggested that near infrared hyperspectral imaging can be used to identify dried jujube varieties quickly.
【作者单位】：浙江大学生物系统工程与食品科学学院;
【基金】：国家科技支撑计划项目(2014BAD04B04)资助
【分类号】：TS255.4;TP391.41

【参考文献】