【作者】 田潇瑜； 彭彦昆； 马世榜； 徐杨； 牛力钊；

【Author】 Tian Xiaoyu,Peng Yankun,Ma Shibang,Xu Yang,Niu Lizhao(College of Engineering,China Agricultural University,Beijing 100083,China)

【机构】 中国农业大学工学院；

【摘要】 嫩度是评价牛肉品质的重要指标。本研究利用可见及近红外区域(VIS/NIR)的反射光谱信息预测牛肉的剪切力值(WBSF),并判别牛肉不同的嫩度等级。以牛肉背最长肌为研究对象,选取宰后成熟14天的样本31个,在牛肉新鲜切口处采集波长范围400～1700nm的反射光谱信息,使用肉类嫩度测量仪测量牛肉剪切力值。利用偏最小二乘回归法(PLSR)分别在可见及近红外(400～1000nm)、近红外(1000～1700nm)以及全波段(400～1700nm)范围内建模,结果表明在1000～1700nm范围内的信息能更好的预测牛肉嫩度,预测相关系数RV达到0.91,预测标准差SEV为3.86N。对比使用逐步回归方法,根据反映牛肉嫩度的特征波段不同组合,建立牛肉品质的多元线性回归(MLR)预测模型,预测相关系数RV达到0.88,预测标准差SEV为8.76N,也能较好的预测牛肉的嫩度。通过预测结果评价牛肉的嫩度等级,准确率最高能达到90%以上,说明利用可见及近红外光谱预测牛肉嫩度是可行的。更多还原

【Abstract】 Tenderness is an important parameter for beef quality judgment.This study used VIS/NIR reflectance spectroscopy to detect Warner-Bratzlar Shear Force(WBSF).31 strip loins samples were aged for 14 days,cooked to an internal temperature of 70 O C.Reflectance spectra range of 200～1700nm were extrcted from fresh cut of beef,and then Warner-Bratzler(WB)shear force were measured.Least-squares regression model(PLSR)was set up in range of 400～1000nm,1000～1700nm and 400～1700nm.The result showed that the model in 1000～1700nm had higher accuracy,with r=0.91.And multi-linear regression(MLR)analysis was used to establish prediction models for tenderness using different wave bands combination,simultaneously.The result showed that the prediction model with feature bands had high accuracy with R=0.89.Experiments showed that,shear force value of model prediction,synthetically evaluated beef tenderness level,of which the classification accuracy reach 90%.更多还原