Abstract:

Two different portable near infrared spectrometers (USB 2000+ and Maya 2000) were used to acquire spectra
of eggs stored for different times with high- and low-intensity light source systems. Discrimination and classification
of storage times of eggs was achieved by two different methods, PCA and LDA analysis. The results showed that LDA
was superior to PCA and Maya 2000 which correctly discriminated 95.7% of egg samples in LDA analysis and 93.2%
of cross-validated samples was superior to USB 2000+. Even using a low-intensity light source system, discrimination
of the freshness and shelf life of eggs can be achieved by an appropriate near infrared spectrometer combined with an
appropriate discrimination method.

Key words: near infrared transmittance spectroscopy, principal component analysis, egg, discrimination analysis