Application of Near-infrared Reflectance Spectroscopy on Prediction of Chemical Composition on Diannan Small Ear Pork Muscle

Near infrared reflectance spectroscopy (NIRS) was used to study the moisture, fat and protein content of Diannan small ear pig intact meat and homogeneous meat modeling to find the best spectral preprocessing methods. Based on 111 pork samples and 11 000~4 300 cm-1, through multiplicative scatter correction (MSC), the second derivative, variable standardization correction (SNV) pretreatment method,using partial least squares (PLS), the near infrared predictive model of Diannan small ear pig pork about moisture, fat, protein was established and the best spectral pretreatment method and the principal component scores were investigated. Moisture was preferably predicted using the MSC of spectra of whole meat processing,R2 was 0.975, RMSEC was 0.304, RMSEP was 1.640; the better predicted fat was homogeneous meat SNV spectral,R2 was 0.911, RMSEC for 0.409, RMSEP was 0.923; the better prediction of protein is the raw homogeneous meat,R2 was 0.946, RMSEC was 0.273, RMSEP was 1.101. Thus, the near-infrared spectroscopy method can detect the fat and water of Diannan small ear pig pork, poor test results of protein.