近红外反射技术开放式检测棉籽中水分和油含量的研究

研究了近红外漫反射分析技术在棉籽，尤其是毛棉籽的水分和油含量检测中的应用。建立了毛棉籽水分和油含量的近红外分析模型，并随机抽样验证了检测模型的检验精度。水分的相关系数为0.965，含油的相关系数0.953，定标标准偏差分别为0.226和0.391。检测结果达到了代替传统检测方法的要求。另外利用部分样品建立了仁含油和仁出率的近红外检验模型，证明近红外方法也可以应用于其检测。

Measurement of Moisture and Oil Content in Gross Cottonseed Based on Near-Infrared Reflectance Technique by Open Detecting Mode

In the present paper,the moisture and oil content in gross cottonseed analyzed by near infrared reflectance spectroscopy(NIRS) technique was studied.Including the basic theory of NIRS and advantage of open detecting mode.The calibrations of the moisture and oil content in gross cottonseed were performed.The precision and accuracy of the moisture and oil content were clarified by random sampling.The correlation coefficient is high: 0.965 for moisture and 0.953 for oil.The calibration errors(SECV) for the moisture and oilcontent in gross cottonseed are 0.226 and 0.391 repectively.The result of NIRS prediction has been proved that the NIRS method can replace the normal analysis method completely.Further studies on the calibration of oil content in nucleus and rate of nucleus in gross cottonseed were carrced out with part of samples.It has been proved that NIRS method can be applied to detect the oil content in nucleus and rate of nucleus in gross cottonseed.