利用近红外谷物分析仪对芝麻整粒进行光谱扫描,分别建立了白芝麻和黑芝麻灰分含量快速检测近红外分析模型,井对光学处理和数学处理过程中不同因素对模型的影响进行了探讨.模型内部验证和外部检验结果表明:去散射方式采用去散射处理,数学处理技术采用“3,4,4,1”(即每间隔4个光谱点进行三阶导数处理和一次平滑处理,不进行二次平滑处理)得到的方程为白芝麻灰分的最佳定标模型,其1-VR值为0.8203,SECV值为0.1837.去散射方式采用标准正常化结合散射处理,数学处理技术采用“2,2,2,1”(即每间隔2个光谱点进行二阶导数处理和一次平滑处理,不进行二次平滑处理)得到的方程为黑芝麻灰分的最佳定标模型,其1-VR值为0.7347,SECV值为0.2334.本文验证了近红外光谱技术快速检测芝麻灰分含量的可行性,表明该分析技术可以应用于芝麻品质的快速评价.%In this paper,we respectively built an NIR analysis model for rapid determination of ash content of white sesame and black sesame by using an FOSS NIR grain analyzer to scan the spectra of whole sesame grains, and discussed the factors influencing the model by adopting optical treatment, mathematic treatment, and so on. The internal and external model validation results showed that the equation obtained by adopting De-trend treatment and a 3,4,4,1 mathematic treatment technique, which included carrying out third derivative treatment and primary smoothing treatment at an interval of each four spectral points and not carrying out secondary smoothing treatment, was the optimum model for calibrating the ash content in white sesame, wherein the 1-VR value was 0.820 3,and the SECV value was 0. 183 7; and the equation obtained by adopting SNV and Detrend treatment and a 2,2,2,1 mathematic treatment technique, which included carrying out third derivative treatment and primary smoothing treatment at an interval of each two spectral points and not carrying out secondary smoothing treatment,was the optimum model for calibrating the ash content in black sesame,wherein the 1-VR value was 0. 734 7 , and the SECV value was 0. 233 4. The paper validated the feasibility of NIR technique in rapid determination of ash content in sesame, and showed that the NIR technique could be applied in the rapid quality evaluation of sesame.
展开▼
