Non-Destructive Detection Of Apple Moldy Core And Watercore By FT-NIR And Electronic Nose

With suspected moldy core,suspected watercore and corresponding varieties of healthy apples as the test material.Explore the near infrared spectroscopy?NIR?and electronic nose technique were combined with chemometrics to detect the apple moldy core and watercore.For the near infrared spectrum and electronic nose technology applied in postharvest apples better internal diseases of discriminant analysis,reasonable processing and utilization to provide theoretical reference.Objective to compare the effects of moldy core and watercore on the physical and chemical quality indexes of the good pulp in the fruit.Provide a reference for consumers to buy healthy fruits.The results are as follows:?1?The near infrared models distinguished moldy core of apple were built.Principal component analysis of near infrared spectra for moldy core apple and healthy apple,2chemometric methods are combined with Fisher discriminant function and MLP neural network,apple moldy core discrimination model is established.The results show that the first15 principal components of 12000 cm^-14000 cm^-11 are combined with Fisher discriminant function and MLP neural network model respectively,and the correct discrimination rates of the validation set are 72%and 87.7%,respectively.Therefore,MLP neural network model can better identify the apple moldy core.It is feasible to use the principal component respectively to combine the Fisher discriminant function and the MLP neural network to distinguish the apple moldy core.However,the accuracy of the identification model needs to be improved.?2?The electronic nose models distinguished moldy core of apple were built.Comparison of electronic nose respectively combined with MLP neural network,RBF neural network and Fisher discriminant three methods to identify apple moldy core.Considering the correct discriminant rate of the model set and the validation set,the MLP neural network model is the best one for the identification of apple moldy core,the correct rate of the training set and the test set is 87.9%and 86.2%,respectively.Among them,the discriminant accuracy of the training set and validation set was 68%and 64.7%,respectively.Therefore,the accuracy of the electronic nose technology combined with chemometric methods for the identification of apple moldy core needs to be further improved.?3?The near infrared models distinguished watercore of apple were built.After 5methods of pretreatment of the near infrared original spectra of the watercore and healthy apple,the principal components of the whole band spectrum were combined with the Fisher discriminant function to modeling.The results show that the near-infrared spectra are pretreated by five methods after MSC,VN,MMN,9 point smoothing and first derivative?9point smoothing?,the Fisher discriminant function modeling is established using the first 20principal components of the 12000 cm^-14000 cm^-11 wavenumber range.The correct discriminant rates for the modeling sets respectively were 93.8%,94.3%,92.3%,94.3%and100%,and the prediction accuracy of the Fisher discriminant function on the unknown sample is 100%by the first order derivative?9 point smoothing?.Because the first derivative?9smooth?pretreatment combined with principal component built by Fisher discriminant model for modeling set and validation set the correct discriminant rate were 100%,so the first derivative?9 point smoothing?model is better than other models,It is verified that the near infrared diffuse reflectance spectroscopy can be used for non-destructive testing of apple watercore.?4?The electronic nose models distinguished watercore of apple were built.The discriminant models of 3 kinds of electronic nose,which are Fisher discriminant,multilayer perceptron neural network and radial basis function neural network,are compared,the correct discrimination rates of unknown samples were 89.7%,89.5%and 85.7%,respectively.Among them,Fisher discriminant analysis and multilayer perceptron neural network have a good effect,but the difference is not obvious.It is verified that the electronic nose technology combined with chemometric methods can be used to identify apple watercore.?5?Study on the difference of physical and chemical quality indexes for good pulp of healthy apple and moldy core apple.With the occurrence of apple moldy core,the physical and chemical properties of the good pulp in the fruit were changed,density and SSC decreased significantly?P?0.05?,the content of total phenolics and flavonoids increased significantly,titratable acid content difference was not significant,to 0.21%;Although the firmness of moldy core apple was 7.82 N/cm² higher than that of healthy apple,the difference was not up to a significant level?P?0.05?.The results showed that the diseased parts of the moldy core apple will affect the physical and chemical quality of the whole fruit,and reduce the nutritional value and health function.For the apple post-harvest timely identification of mold disease to provide physical and chemical quality reference,it is recommended that consumers should not eat the good pulp of moldy core apple.?6?Study on the difference of physical and chemical quality indexes for good pulp of healthy apple and watercore apple.With the occurrence of heart disease,the physical and chemical quality of apple's good pulp changed.With the occurrence of heart disease,the physical and chemical quality of apple's good pulp changed,the titratable acid content decreased significantly,Density,hardness,SSC and flavonoid content increased significantly, total phenolics were higher than those in healthy apple,but the difference was not significant?P?0.05?.Indicating that watercore can influence the physical quality of the whole fruit,but the slight watercore does not affect the fruit to eat,because of the taste of people loved,instead of apple postharvest timely identify waterccore provides the physical and chemical quality reference.