Pork and beef classification method based on Raman spectrum

Abstract

The invention discloses a pork and beef classification method based on a Raman spectrum. Raman spectrum data on the surfaces of pork and beef samples is collected through a confocal microscopic Ramanspectrum instrument; the Raman spectrum data is subjected to preprocessing operation; then, principal component analysis is performed to screen out five Raman feature spectrum peaks; and next, the results are substituted into a classifier for classification. The pork and beef classification method based on Raman spectrum has the advantages that high accuracy is realized on the classification on the pork and the beef; the classification speed is high; and the robustness is high.

Description

technical field [0001] The present invention relates to a Raman spectrum meat classification method, in particular to a Raman spectrum-based pork and beef classification method, especially five Raman characteristic spectrum peaks that are finally used for beef and pork classification models . Background technique [0002] Meat is one of the important sources of nutrients for the human body. In recent years, food safety problems have emerged in various countries. Some illegal traders use meat of lower economic value as adulterants to mix meat products of high economic value in order to seek economic benefits. This kind of "economic interest-driven adulteration" has seriously damaged the interests of consumers and suppliers, and shaken the public's confidence in the safety of meat products. [0003] At present, the identification methods of meat products at home and abroad are mainly based on physical and chemical analysis, including enzyme-linked immunosorbent assay (ELISA...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problems existing in the background technology, in order to solve the technical problem that the original Raman spectral data has a lot of fluorescence signal interference and the spectral peaks are dense, which will cause the classification to reduce the accuracy of the model and increase the difficulty of classification, the present invention analyzes the collected Raman spectrum data of minced beef tenderloin and pork tenderloin, a method of pork and beef classification based on Raman spectroscopy is proposed, and the Raman spectrum peak that best characterizes the characteristics of the sample is found, which greatly improves the robustness of the classification process , to speed up spectral data processing

Images