Research On Nondestructive Detection Of Baby Cabbage Freshness Based On Electronic Nose And Near Infrared Spectroscopy

The traditional quality test of baby cabbage is in laboratory,which requires the destruction of baby cabbage samples.At the same time,the quick detection of freshness and quality index of baby cabbage is one of the hot spots that the market and consumers pay close attention to.Under the support of Jiangsu Province Agricultural Science and Technology Independent Innovation Fund Project "High-grade Vegetables Post-harvest Commodity Rate Maintenance Technology Innovation and Integrated Application"(2018-2021,Project No.CX(18)2028),this study mainly uses electronic nose technology and near infrared spectroscopy technology to carry out non-destructive testing on the quality of baby cabbage during the storage,respectively from color,weight loss rate,hardness,Vc and other physical and chemical indexes.In addition,through the data import,modeling,analysis to nondestructively detect its freshness.The specific research contents and results are as follows:1.Effect of different packaging materials on main quality indicators of baby cabbage during storageTaking Brassica campestris(baby cabbage)as the material,the fresh baby cabbage was packed with four different packaging bags(polyethylene bags,polyvinyl chloride bags,antifog bags and nano-packaging bags)and then stored under low temperature(4℃)and ormal temperature(20℃)conditions.The indicators of the weight loss rate,color value(L*,a*,b*),chlorophyll,Vc,hardness were investigated to study the changes of quality and physiology of baby cabbage and explore the optimum packaging material.The results showed that all these four packaging treatments significantly maintained the quality of baby cabbage,mainly by reducing the weight loss,keeping the color,maintaining the hardness,inhibiting the reduction of the Vc and chlorophyll content,further delaying the aging and browning of baby cabbage.Besides,baby cabbage packaged with nano-packaging bags exhibited better quality than those ones packaged with three other materials.The mass loss rate of the nano-package group was only 0.59%when stored for 15 d under(20±0.5)℃,while the control group reached 55.8%,the difference was significant(P<0.05).L*value,a*value and b*values were higher than the control group.The hardness of the nano-package group was maintained at 20.0 N,the control group decreased to 12.3 N from the initial hardness 26.0 N(P<0.05).The contents of Vc and chlorophyll were 45.0 mg/g and 0.025 mg/g,respectively,which were significantly higher than those of the control group.The baby cabbage was stored in a storage environment of(4±0.5)℃ and relative humidity of 85%～95%,which was similar to the normal temperature.The nano-packging bags group was significantly better than that of the control group and exhibited good commodity values for 30 d of storage.2.Study on quality prediction of baby cabbage based on near infrared spectroscopyIn this study,the near infrared spectrum information of baby cabbage was obtained under the condition of 20℃ and 85-95%relative humidity.Based on full-band spectrum,partial least squares regression(PLS)and support vector machine(SVM)regression were used on the quantitative modeling of mass loss rate,L*a*b*,hardness,and Vc content.And the freshness of baby cabbage was classified based on the change of quality index under room temperature storage.The results show that for weight loss,the full-band PLS-SNV prediction model works best,with Rp2 and RMSEP of 0.96 and 1.432,respectively;for the L*value,the full-band SVM-MSC prediction model works best,and Rp2 and RMSEP are 0.82 and 2.013;for b*values,the full-band PLS-1-st prediction model performs relatively well,with Rp2 and RMSEP being 0.85 and 1.264,respectively;for hardness,the full-bandSVM-Autoscale prediction model performs relatively well,with Rp2 and RMSEP,as 0.60,2.453,respectively.For Vc content,the full-band PLS-MSC prediction model performs relatively well,with Rp2 and RMSEP of 0.95 and 3.192,respectively.3.Detection of odor changes in baby vegetables based on electronic nose and GC/MSThe PEN3 electronic nose was used to extract and store the odor values of baby food for 0,3,6,9,12,and 15 days.Finally,the optimized arrays of the sensors were determined as S6,S7,and S9.Based on the PLS-DA and SVM-C algorithms,a discriminative model for the freshness of baby cabbage under normal temperature storage was established.The overall discriminant accuracy of the validation set can reach 88.8%and 92.2%,respectively.GC-MS was used to detect the volatiles during storage of baby cabbage,and the research found that nitrile compounds,ester compounds such as phenylpropionitrile,benzobutyronitrile,benzonitrile,5-methylhexanitrile,1H-indole-3-acetonitrile,etc.1-butenyl-4-isothiocyanate,hexyl isothiocyanate,phenethyl isothiocyanate,alcohol compounds 2-hexene-l-ol,2-ethyl-1-hexane Alcohol,phenethyl alcohol,3,7-dimethyl-1,6-octadien-3-ol,ether compounds dimethyl disulfide and dimethyl trisulfide are characteristic volatile substances.These compounds correspond to the screened sensors and provide theoretical support for sensor optimization.Compared with the control group,the nano-packaging group had a higher content of 2-hexenal,a lower content of alcohols,and more types and contents of esters and nitriles than the ck group.4.Research on rapid detection of baby cabbage freshness combined with electronic nose and near infrared spectroscopyTaking baby cabbage as a research object,the electronic nose and near-infrared spectrometer were used to determine the odor value and spectral information of the same batch at the same storage time point(0,3,6,9,12,15 d).After preprocessing the data by the normalization method,the model of the freshness of baby cabbage was established based on PLS-DA and SVM-C algorithms.The results show that the overall discrimination accuracy of its validation set can reach 94.0 and 93.7%,respectively.The modeling effect is better than single detection technology,with higher detection accuracy and better prediction effect.Among them,for the prediction of the quality index of baby cabbage,the prediction models of each index have improved significantly,especially the hardness.