The Identification Of Species Blood Based On Raman Spectroscopy

In recent years,with the continuous progress and development in the field of biopharmaceuticals,the exchange of blood products in the market and the exchange of immigration have become increasingly frequent.The identification of blood species is also increasingly important.The traditional methods all have more or less problem,and it is impossible to achieve rapid detection and identification of blood sample species.This article focuses on Raman spectroscopy experiments for the identification of human and multi-animal blood,using micro Raman spectroscopy.The excitation wavelength screening experiment of the blood sample container and the excitation light source before the experiment was performed by Renishaw confocal Raman spectroscopy,and Raman spectroscopy experiments of human,canine,pig and chicken blood were conducted by a Raman spectroscopy experimental system.the study.The data obtained from the experiment was used to study the spectral pretreatment method.Based on the analysis of experimental data and noise signal principle,db5 wavelet transform was used to perform 3-layer filter denoising and minimum piecewise polynomial fitting baseline correction to achieve spectral preprocessing of blood spectral experimental data.By studying the experimental data classification algorithm,the k-neighborhood algorithm classifier and the random forest algorithm classifier are selected for comparison.Finally,according to the 100% discrimination rate of the random forest algorithm classifier and the trainer model accuracy score of 0.9167,the ratio k is determined.The algorithm classifier is more suitable for the identification of Raman spectra of human,dog,pig,and chicken blood.