近红外光谱对特征部位牛肉的分析

Prediction of Chemical Parameters of Different Cuts of Beef by Near Infrared Transmission Spectroscopy

通过采集牛霖、牛柳、牛腩3个部位牛肉样品的近红外光谱并对其蛋白质、脂肪、水分和灰分等化学指标进行测定,应用偏最小二乘法分别建立蛋白质、脂肪、水分和灰分的校正模型。牛霖样品的灰分和水分校正模型相关系数分别为0.9843、0.9740,预测相关系数分别为0.9502、0.9342;牛柳样品的蛋白质和水分校正模型相关系数分别是0.9540、0.7752,预测相关系数分别为0.9500、0.9060;牛腩样品的灰分和脂肪校正模型相关系数分别是0.9746、0.9218,预测相关系数分别是0.9560、0.9233;牛霖和牛柳样品的水分总校正模型相关系数为0.7504,预测相关系数是0.9737;牛腩和牛霖样品的灰分总校正模型相关系数是0.8770,预测相关系数为0.9777。结果表明:近红外光谱分析法预测准确度较高,可以用于评价牛肉品质。

Beef thick flank,fillet and brisket were scanned using a near infrared spectrophotometer and their chemical parameters such as protein,fat,water and ash were measured.A calibration model for each of these parameters was set up by means of partial least squares（PLS） regression.The correlation coefficients of calibration for ash and water contents in beef thick flank were 0.9843 and 0.9740,respectively,and the correlation coefficients of validation were 0.9502 and 0.9342,respectively.The correlation coefficients of calibration for protein and water contents in beef fillet were 0.9540 and 0.7752,respectively,and the correlation coefficients of validation were 0.9500 and 0.9060,respectively.The correlation coefficients of calibration for ash and fat contents in beef brisket were 0.9746 and 0.9218,and the correlation coefficients of validation were 0.9560 and 0.9233,respectively.The correlation coefficients of calibration and validation for total water content in beef thick flank and fillet was 0.7504 and 0.9737,respectively,while those for total ash content in beef brisket and thick flank were 0.8770 and 0.9777,respectively.This study shows that near infrared spectroscopy allows accurate predictions of chemical parameters of beef and therefore can be used to evaluate beef quality.