Quality Detection And Safety Discrimination Of Pork With Near-infrared Spectroscopy

The pork is the main diet of the Chinese consumers, and its consumption is much largerthan the other livestock and poultry meat. In the recent years, some unscrupulous traders sellstale or putrid meat to consumers in order to make a large profit, these behaviors damage theconsumers’ rights and health. It is necessary to take an effective, simple and feasible methodto detect and control the pork freshness and quality in processing, transportation, storage,marketing and other aspects.In this study, near-infrared spectroscopy was used to detect pork without damage, andthen the chemometrics models about pork cooking loss, tenderness, pH value, total volatilebasic nitrogen content(TVB-N) and total number of bacterial colony were established foranalysis. These models can provide testing basis for detecting pork quality and safety. Themain results are shown as follows:(1) The near-infrared spectroscopy was used to determine pork cooking loss, tenderness,pH value, total volatile basic nitrogen content and total number of bacterial colony these fivequality indices, and then the quantitative analysis models and the qualitative discriminantmodels which have high correlation coefficient, determination coefficient and predictionaccuracy were established respectively. Thus, the detection and prediction about pork qualityindices content have feasibility with near-infrared spectroscopy;(2) The cluster analysis about near-infrared spectrum combined with sensory evaluationof pork was applied comprehensively to judge meat freshness, which stored at roomtemperature and refrigeration temperature. Then, pork pH value and TVB-N content thresholdat these three kinds of freshness were classified. At different storage temperatures, pork pHvalue and TVB-N content had different change rate and different variation range;(3) The multilayer perceptron and radial basis function artificial neural network wereapplied to establish qualitative discriminant models about pork pH value and TVB-N content,which were used to discriminate and judge the pork freshness. These models verified theaccuracy of freshness threshold about pork pH value and TVB-N content with high freshnessdiscrimination accuracy rate;(4) The zero level mode of the dynamic model was used to establish the kinetic model oftotal number of bacterial colony and the regression model was established based on thespectral eigenvalues and total number of bacterial colony with principal component analysis.The correlation coefficient of these models reached more than0.9. Thus, the detection andprediction of pork storage and shelf life with near-infrared spectroscopy were feasible.In this research, near-infrared spectroscopy and chemometric methods were applied todetect the pork quality indices. The quantitative discriminant model, qualitative discriminantmodel and kinetic model were established based on near-infrared spectra and indices data, and then were used to discuss the determination about pork quality indices content, thediscrimination about pork freshness and the prediction about pork storage and shelf life withfeasibility. These can provide effective detection methods to distinguish pork freshness andprovide the theoretical basis and technical support for detecting the pork quality and safety.