基于高光谱成像技术的梅河大米产地确证因子研究

为探究吉林省梅河大米的产地确证因子,本研究采用空间网格布点法对吉林省梅河口市地理标志大米进行样本采集。通过梅河大米的高光谱信息和蛋白质、直链淀粉、棕榈酸、油酸、亚油酸、脂肪酸含量进行光谱特征波长提取、相关性分析和产地判别。结果表明,梅河大米的理化指标具有明显的产地特征,高光谱九个特征波长TB_1～TB_9(455、515、583、615、621、776、874、923、942 nm)与上述六个指标呈现显著相关性。利用高光谱九个特征波长对梅河大米与柳河县大米进行SPSS判别分析,整体正确率为95%。SIMCA偏最小二乘法建立的模型可以将梅河大米样品与柳河县大米样品进行区分。说明梅河大米高光谱九个特征波长可作为其产地确证因子。

Study on Confirmation Factors of Meihe Rice Producing Area Based on Hyperspectral Imaging Technology

In order to explore the confirmation factors of the Meihe rice production in Jilin Province, this paper uses the spatial grid layout method to collect samples of the geographical indication rice in Meihekou City, Jilin Province. Through the hyperspectral information of Meihe rice and protein, amylose, palmitic acid, oleic acid, linoleic acid and fatty acid content, spectral characteristic wavelength extraction, correlation analysis and origin discrimination were carried out. The results showed that the physical and chemical indicators of Meihe rice had obvious characteristics of origin. The nine characteristic wavelengths of high-spectrum TB1~TB9 (455nm, 515nm, 583nm, 615nm, 621nm, 776nm, 874nm, 923nm, 942nm) were significantly correlated with the above six indicators. The SPSS discriminant analysis of Meihe rice and Liuhe county rice was carried out by using nine characteristic wavelengths of hyperspectral, and the overall correct rate was 95%. The model established by SIMCA partial least squares method can distinguish the Meihe rice samples from the Liuhe County rice samples. It is indicated that the nine characteristic bands of Meihe rice hyperspectral can be used as the confirmation factor of its origin.