冻藏猪肉在近红外光谱应用中的快速无损检测

Rapid nondestructive testing of frozen pork by near-infrared spectroscopy

利用可见-近红外光谱技术对新鲜肉和长期冻藏猪肉水分含量进行定量模型的建立。采集冻藏0、3月的新鲜肉和冻藏6、9月的长期冻藏肉的光谱图,分别与其水分含量值一一对应,选用偏最小二乘回归结合二阶微分处理方法建立定量判别模型,所得新鲜肉和长期冻藏猪肉定量模型的相关系数分别为0.8714、0.9614,校正均方根误差分别为1.10、0.3169。所建模型可较好的判别新鲜肉与长期冻藏猪肉。

A quantitative model for the moisture content of fresh meat and long-term frozen pork was established by visible-near infrared spectroscopy. The spectral image of fresh meat frozen for 0 and 3 months and long-term frozen meat frozen for 6 and 9 months were collected and corresponded to their moisture content values one by one. Partial least squares regression was selected and combined with a second-order differential processing method, and a quantitative discriminant model was established. The correlation coefficients of the fresh meat and long-term frozen pork quantitative models were 0.8714 and 0.9614 respectively, and the corrected root mean square errors were 1.10 and 0.3169 respectively. The established model could better distinguish between fresh meat and long-term frozen pork.