Abstract
Macadamia kernels are visually sorted based on the presence of quality defects by specialized labors. However, this process is not as accurate as non-destructive methods such as near infrared spectroscopy (NIRS) and nuclear magnetic resonance (NMR). Thus, NIRS and NMR in combination with chemometrics have become established non-destructive method for rapid assessment of quality parameters in the food and agricultural sectors. Therefore, the quality of macadamia kernel was assessed by NIPS and NMR using chemometric tools such as PCA-LDA and GA-LDA to evaluate external kernel defects. Macadamia kernels were classified as: 1 = good, marketable kernels without defects; 2 = kernels with discoloration; 3 = immature kernels; 4 = kernels affected by mold; and 5 = kernels with insect damage. Using NIRS, the GA-LDA resulted in an accuracy and specificity of 97.8% and 100%, respectively, to classify good kernels. On the other hand, PCA-LDA technique resulting in an accuracy higher than 68% and specificity of 97.2% to classify immature kernels. For NMR, PCA-LDA resulted in an accuracy higher than 83% and GA-LDA resulted in an accuracy of 100%, both to classify kernels with insect damage. NIRS and NMR spectroscopy can be successfully used to classify unshelled macadamia kernels based on the defects. However, NIRS out-performed NMR based on the higher accuracy results.