Rapid Evaluation Of Lateolabrax Japonicus Freshness Based On Near Infrared Spectroscopy

Aquatic products are liable to suffer spoilage after fishery harvesting because of its abundant nutrition and high moisture,which gives normal trading some big challenges. The definition of freshness as a function of storage time is an effective method of shelf life estimation.The shorter the storage time, the better its original characteristics. Therefore, freshness makes a major contribution to the quality of sea foods. At present, there are many methods of sea foods evaluation. However, these methods are all time-consuming, destrusive and operation-complex, they can’t be applied in food analysis. As a pollution-free and efficient analysis technology, near infrared spectroscopy has huge potential for assessing the freshness of aquatic products rapidly and objectively.In view of this, the study was based on establishment of storage time evaluation of lateolabrax japonicus, then rapidly estimating the quality of lateolabrax japonicus combined with partial least squares and characteristic wavelengths selection, differentiating fresh and frozen-thawed fish by SIMCA and PCA-DA, and discussing the differentiation theory initially, simultaneously studying the effect of refrigeration after freeze on lateolabrax japonic us. The results are showed as following:1. NIR spectra in the range of1000-2500nm were collected for qualitative and quantitative analysis. Near infrared spectroscopy combined with PCA and SIMCA, established qualitative analysis model of differernt parts, all of the prediction accuracy reached100%. The model between spectra of lateolabrax japonicus and storage time was established and developed by multivariate analysis software, the best model from skin with1000-1800nm can obtain correlation coefficient of0.95and prediction error of1.71d.2. By determining some physical and chemical indicators of lateolabrax japonicus, established quantitative analysis model, compared the effects of different pretreatment on models, acquired the optimal TVB-N model, hardness model and L*model, the correlation coefficient and root mean squared error of calibration were0.88814,3.35,0.83451,31.8,0.80443,0.672, respectively and the coefficient and root mean squared error of prediction were0.83076,3.59;0.75968,34.5;0.75968,34.5, respectively. Meanwhile, TVB-N model was improved by interval PLS and synergy interval PLS, synergy interval PLS was better than interval PLS, the correlation coefficient and root mean squared error of calibration and prediction were respectively0.98336,2.75;0.92679,3.1.3. By using SIMCA and PCA-DA, fresh and frozen-thawed fish were differentiated effectively in1000～1100nm, and the result displayed that PCA-DA discriminant analysis was superior to SIMCA, recognition rate was all over95%. The changes of tissue micro structure and protein from lateolabrax japonic us showed that gap size was the basis of differentiation, and myofibrillar protein also had a certain relation with spectra of fish. With the change of quality of frozen-thawed fish measured, all the parameters had effects on quality except filtration rate. The muscle tissue microstructure changes exhibited that the damage caused by freeze and thaw was obvious and irreversible.