Hyperspectral Model And Spatial Distribution Of Myoglobin Content In Cooked Meat Of Tan Sheep During Storage Period

Meat color is regarded as a Visual measure of the freshness of lamb products,its chemical essence is myoglobin molecules.The three forms of myoglobin in muscle tissue are deoxymyoglobin(DeoMb),oxymyoglobin(OxyMb)and methemoglobin(MetMb).Meat color is determined by the molecule structure and proportion of DeoMb,OxyMb and MetMb and their mutual transformation reaction There are few reports on the relationship between spectral information of cooked meat color and spatial distribution of myoglobin content in Tan sheep during storage based on Vis/NIR and NIR hyperspectral imaging technology,the color changes of cooked meat of Tan mutton during storage period were analyzed with myoglobin(DeoMb,OxyMb and MetMb)as detection indexes,and the best model was selected to explore the spatial distribution of myoglobin content of cooked meat of Tan mutton during storage period.The primary conclusions are as follows:(1)The color of Tan mutton was predicted by Vis/NIR hyperspectral imaging technology combined with different pretreatment methods,feature wavelength extraction algorithms and model development methods.The results showed that the normalized pretreatment CARS-PLSR model was better than FS-PLSR model in predicting DeoMb(RC=0.850.RP=0.868)under refrigeration conditions,and the median filter pretreatment CARS-PLSR model was better than FS-PLSR model in predicting OxyMb(RC=0.835,RP=0.844).The smoothing SGolay pretreatment CARS-PLSR model was the best in predicting MetMb(RC=0.846,RP=0.836).Under storage conditions of room temperature,the CARS-PLSR model pretreated by smoothing SGolay was better than FS-PLSR model in predicting DeoMb(RC=0.932,RP=0.957),and the CARS-PLSR model pretreated by gaussian filter was better than FS-PLSR model in predicting OxyMb(RC=0.924,RP=0.903).In addition,the CARS-PLSR model pretreated by SNV was the best in predicting MetMb(RC=0.772,RP=0.798).(2)The prediction model of cooked Tan mutton color was established by NIR hyperspectral imaging technology,optimized the best pretreatment method,feature wavelength extraction algorithm and model construction method.The results showed that the normalized pretreatment CARS-PLSR model was better than FS-PLSR model in predicting DeoMb(RC=0.758,RP=0.713).The CARS-PLSR model pretreated by deresolve was better than FS-PLSR model in predicting OxyMb(RC=0.873,RP=0.807),while the CARS-PLSR model without pretreatment was better than FS-PLSR model in predicting MetMb(RC=0.869,RP=0.864).Under storage conditions of room temperature,the CARS-PLSR model pretreated by smoothing SGolay had a better effect in predicting DeoMb(RC=0.780,RP=0.888)and the CARS-PLSR model pretreated by detrend had a better effect in predicting OxyMb(RC=0.763,RP=0.737).while the CARS-PLSR model pretreated by gaussian filter had a better effect in predicting MetMb(RC=0.860,RP=0.875).(3)The spatial distribution map of myoglobin content was established in cooked meat of Tan sheep during storage period.The results showed that the content of DeoMb decreased gradually from inside to outside with storage time,the content of OxyMb increased first and then decreased in the middle part of samples with storage time,and the content of MetMb increased gradually from inside to outside with storage time.