Method for discriminating urea-doped milk based on two-dimensional correlation near-infrared spectrum gray-level co-occurrence matrix

Abstract

The invention discloses a method for discriminating urea-doped milk based on a two-dimensional correlation near-infrared spectrum gray-level co-occurrence matrix. The method comprises: obtaining the two-dimensional correlation near-infrared spectrums of pure milk and different concentrations of urea-doped milk as training samples, carrying out gray-level quantization on the two-dimensional correlation near-infrared spectrums of the training samples, calculating the means and the standard deviations of the gray-level co-occurrence matrix characteristic parameters, using the means and the standard deviations as the texture characteristic parameters of the two-dimensional correlation near-infrared spectrums, carrying out normalization treatment on the texture characteristic parameters to obtain normalized texture characteristic parameters, establishing a support vector machine model, respectively substituting the eight normalized texture characteristic parameters of each training sample into the support vector machine model to obtain a trained support vector machine model, and substituting the normalized texture characteristic parameters of a to-be-tested sample into the trained support vector machine model to determine whether the to-be-tested sample is doped. The method of the present invention has high discrimination efficiency, and can quickly identify a large amount of urea-doped milk.

Description

technical field [0001] The invention belongs to the technical field of milk detection, and specifically relates to a method for discriminating urea-doped milk based on a two-dimensional correlation near-infrared spectrum gray-scale co-occurrence matrix. Background technique [0002] Milk is known as "white blood" and is the oldest natural beverage. It contains a lot of protein, fat, lactose, inorganic salts and calcium, etc., and is a product with high nutritional value. However, the "melamine" and "leather milk" incidents in recent years have caused a serious crisis to the reputation of the entire dairy industry, and the sales volume in the domestic market has fallen sharply, which has become a key issue restricting the further development of the dairy industry. [0003] After extracting the characteristic information of adulterants in milk, due to the complex variability of the milk system itself, the traces of the adulterants, and the overlapping of characteristic peaks o...

AI Technical Summary

Problems solved by technology

However, the "melamine" and "leather milk" incidents in recent years have caused a serious crisis to the reputation of the entire dairy industry, and the sales volume in the domestic market has fallen sharply, which has become a key issue restricting the further development of the dairy industry

[0003] After extracting the characteristic information of adulterants in milk, due to the complex variability of the milk system itself, the traces of the adulterants, and the overlapping of characteristic peaks of the adulterants and milk-specific components, it is difficult to use one-dimensional near-infrared spectroscopy. Analyze its characteristic peaks

It is also difficult to qualitatively analyze whether there is adulteration in milk by traditional manual comparison of two-dimensional correlation spectroscopy, which has the disadvantages of subjective misjudgment and the inability to achieve a large number of spectral comparisons

Images