Table 2. Recent studies on meat quality detection using near-infrared spectroscopy

Category Measured attribute Analytical method Performance References
Chicken Identification and classification (moisture, lipid contents, protein contents, water holding capacity, and shear force) SVM Accuracy of 91.8% (Geronimo et al., 2019)
Pork Freshness BP-AdaBoost Correlation coefficient of 0.8325 (Huang et al., 2015)
Chicken Water-holding capacity PCA and PLSR Correlation coefficient of 0.91 (Barbin et al., 2015)
Mutton Discriminating the adulteration SVM Accuracy of 90.38%–99.07% (Zhang et al., 2015a)
Pork Moisture PLSR Correlation coefficient of 0.906 (Peng et al., 2018)
Chicken breast Protein LDA and PLSR Accuracy of 99.5%–100% (Wold et al., 2017)
Fish Microbial spoilage PLSR and LS-SVM Correlation coefficient of 0.93 (Cheng et al., 2015)
Rhubarb Identification PLS-DA, SIMCA, SVM and ANN Accuracy of 94.12% (Sun et al., 2017)
Beef Adulteration AF Correlation coefficient of 0.91 (Chen et al., 2018)
Beef, chicken and lard Authentication and classification SVM Accuracy of 98.33% (Alfar et al., 2016)
Turkey meat Identification PLS-DA Correlation coefficient >0.884 (Alamprese et al., 2016)
SVM, support vector machine; BP-AdaBoost, namely back propagation artificial neural network (BP-ANN) and adaptive boosting (AdaBoost); PCA, principal component analysis; PLSR, partial least squares regression; LDA, linear discriminant analysis; LS-SVM, least square support vector machine; PLS-DA, partial least squares-discriminant analysis; SIMCA, soft independent modeling of class analogies; LS-SVM, least square support vector machines; ANN, artificial neural network; AF, artificial fish swarm algorithm.