基于支持向量机的复烤片烟近红外分类模型的建立
Establishment of near-infrared classification model based on support vector machine for redried strips
-
摘要:
为建立复烤片烟的支持向量机（Support Vector Machine， SVM）分类模型，以云南、福建复烤片烟为研究对象，采用近红外光谱仪在4 000~10 000 cm-1 波长范围对片烟样品进行光谱采集，获得复烤片烟样本的基础光谱数据。在有监督学习模式下，采用SVM算法，经一阶导数+SG平滑预处理后，建立了复烤片烟的PCA-SVM分类模型，该模型训练速度较快，对云南、福建复烤片烟的识别准确率分别达到了98.97%和99.59%，取得了满意的分类结果。说明复烤片烟SVM分类模型，可用作片烟质量稳定性评估工具，可快速准确识别出复烤片烟的具体规格。 Abstract:In order to establish a support vector machine (SVM) classification model for redried strips, redried strips from Yunnan and Fujian Provinces were studied. The near-infrared spectra of strip samples were collected in the wavelength range of 4 000-10 000 cm-1, and the basic spectral data of redried strip samples were obtained. Under the supervised learning mode, the PCA-SVM classification model of redried strips was established through SVM algorithm after the first derivative +SG smoothing pretreatment. The training speed of the model was fast, the recognition accuracy rate of redried strips from Yunnan and Fujian reached 98.97% and 99.59% respectively, and the classification results were satisfying. The SVM classification model for redried strips can be used as an evaluation tool for quality stability of redried strips which can quickly and accurately identify the specifications of redried strips.
下载: