Qualitative And Quantitative Analysis Of Milk Based On Hyperspectral

Milk is the most common dairy product in our daily life.It contains a large amount of protein and amino acids required by the human body.Therefore,milk quality evaluation has always been one of the hot topics in the dairy industry.In order to ensure the quality of milk,the country has introduced A series of dairy product inspection standards have been established,but traditional inspection methods are not only time consuming and laborious,but also easily cause environmental pollution.Therefore,this study uses hyperspectral imaging technology and line-scanning Raman hyperspectral imaging system to perform qualitative and quantitative analysis of various complex components in milk.Combined with stoichiometric methods,a rapid milk qualitative and quantitative analysis model is established.Milk composition testing and analysis provides an effective way.This paper mainly researches on the basis of hyperspectral technology and Raman spectral system technology.Taking milk as the research object,a set of Raman hyperspectral imaging detection system is constructed to conduct a qualitative and quantitative analysis of the content of different components in milk.The optical properties of Raman hyperspectral in milk and the future application prospects of hyperspectral technology in milk detection.Specifically,the research content of this article is as follows:(1)Construct a hyperspectral detection system.Use the existing line scan Raman hyperspectral system and set parameters for the system and design the hardware and software of the detection system.The design of the software part mainly focuses on functions Sexual design,the main purpose is to ensure the good operation of the system functions.For the hyperspectral correction,polystyrene material was used in the research process.According to the influence of different parameters on the milk sample,the most suitable parameters for milk detection were selected.(2)Study on Optical Properties of Hyperspectral Signals in Milk Detection.In the courseof this study,a double-layer sample was used to focus on the optical properties of five added ingredients in three different types of milk.The results of the study found that NaSCN(sodium thiocyanate),C3H6N6(melamine),C2H4N4(Dicyandiamine),(NH4)2SO4(ammonium sulfate),CH4N2O(urea),etc.as the milk concentration increases,the signal intensity of the hyperspectral signal in milk is continuously decreasing,and the signal attenuation of the three types of milk remains highly similar.(3)Qualitative analysis of milk based on hyperspectrum.According to the difference in fat content in milk,a hyperspectral system was used to collect multi-dimensional diffuse reflection hyperspectral signals and acquisition of transmission signals of different types of milk and qualitative analysis was performed on the complex components in skim milk and pure milk,.The test uses devices such as laser light sources,CCDs,and monochromators.The main purpose is to collect diffuse reflectance spectral images in biological tissue simulation fluid(Intralipid-10%),and use methods such as Monte Carlo and diffuse approximation methods The backward derivation further proves the science and reliability of the hyperspectral system in this experiment.(4)Quantitative analysis of milk based on hyperspectrum.In the experiment,A total of49 mixed milk samples were selected and divided into 32 correction sets and 17 test sets.A convolutional neural network model was constructed.Through continuous adjustment of the parameters of the convolutional neural network model,the number of convolution kernels and the convolution sum were studied.The effect of size on the generalization performance of the model.Finally,compare the generalization performance of the convolutional neural network model with the traditional machine learning model,and finally determine the superiority of the convolutional neural network prediction ability.