<i>Comprehensive evaluation method of ham spoilage based on hyperspectral technique</i>

In order to solve the problems of single evaluation method of meat products during storage, based on visible/near infrared hyperspectral technique, the main indicators of ham spoilage during storage were studied at 4℃. 92 ham hyperspectral data were collected, and total viable count (TVC), volatile base nitrogen (TVB-N), Color (L, a*, b*) and pH value of ham were measured at the same time. In order to improve the evaluation method of ham spoilage, principal component method was used to extract principal component of index. Variance contribution rate was used as weight to construct comprehensive index function. Then the spectra were processed by different methods, and a partial least squares prediction model was established. The best results showed that the correlation coefficients Rc and Rp were 0.96 and 0.93, and the prediction errors SEC and SEP were 0.36% and 0.60% respectively. Classification of ham spoilage was established by combining CI with TVC. The hams were divided into three levels: fresh, sub-fresh and spoilage. Then the CI and the TVC discrimination model were established by using partial least squares discriminant method. The results showed that the classification model of partial least squares discriminant model had better effect. Identification rates of TVC correction set and verification set were 94.20% and 86.96%, the identification rates of the sensory comprehensive index correction set and verification set were 89.96% and 82.61% respectively. The research showed that the establishment of CI combined with TVC realized quality detection of ham spoilage based on visible/near infrared hyperspectral technology better.