Fast classification of rice (Oryza sativa) cultivars based on fragrance and environmental origins by means of near infrared spectroscopy

Abstract

Abstract Presented paper aimed to investigate the application of near infrared spectroscopy as a fast and robust approach in distinguishing two different rice cultivars based on their fragrance character. Spectra data of rice samples were acquired and recorded in wavenumbers range of 4000-10 000 cm−1 with co-added of 32 scans and 4x optical gain. Further, multivariate analysis in form of principal component analysis (PCA) was employed to classify those samples based on near infrared optical properties. The result showed that both rice cultivars can be precisely clustered and discriminated with total explained variance of 100% using first two principal components (PC). The respective chemical parameters of these differences are fiber content, carbohydrates, amylose content and fragrance. The mentioned three parameters were vibrated strongly in wavenumbers 4260, 7512 and 7900 cm−1 for carbohydrates, amylose and fragrance, while for fiber content, it vibrated at 5183 cm−1 respectively. Based on obtained results, it may conclude that near infrared spectroscopy approach is able to classify and discriminate rice cultivars based on their fragrance by means of spectral optic properties in near infrared region.