บทคัดย่องานวิจัย

หมวดหมู่ : ฐานข้อมูลงานวิจัย

Visible-NIR spectroscopy as a tool for varietal discrimination and characterisation of the harvest stage in apples

Bertrand D. and Laurens F.

5th International Postharvest Symposium . Volume of Abstract . Verona, Italy 6-11 June 2004, p.116

2004

บทคัดย่อ

Visible-NIR spectroscopy as a tool for varietal discrimination and characterisation of the harvest stage in apples   The objective of the study was to evaluate the interest of Visible-NIR spectroscopy as a fast method for the non-destructive assessment of the degree of maturity of apples. The sample collection consisted in fruits belonging to 6 commercial varieties (Gala, Elton, Golden, Jonagold, Braeburn and Fuji) collected at 3 degrees of maturity. For each condition (variety x degree of maturity), 30 apples were studied by Visible-NIR spectroscopy. The spectra were recorded in duplicate o­n a Visible-NIR spectrometer (NIRS system) fitted with a fibre optic probe. The spectra were acquired at wavelength ranging from 400 nm to 2200 nm at 2 nm intervals, leading to 900 data points.

The raw spectral data were processed by principal component analysis (PCA) and stepwise discriminant analysis (SDA), in order to classify the fruits according to their varieties or their maturity stages. SDA were achieved using a cross validation procedure. In 8 independent cross-validation tests, more than 90% of the fruits were correctly classified in their actual varieties. The spectral absorption bands which played a major role in the discrimination were detected at 454, 537, 669 and 1925 nm. In the same way, SDA was applied o­n the whole collection in order to discriminate the degree of maturity of the fruits. In cross validation tests, about 260 fruits among 300 were correctly classified according to their actual maturity stage but the accuracy of the discrimination varied from o­ne variety to another.

In further study, it would therefore be interesting to perform SDA separately o­n each variety. The study showed the interest of NIR spectroscopy as a rapid method for the characterisation of the ripeness of apples. Further data processing will be achieved o­n the sole NIR range, in order to avoid using chlorophyll as a marker of maturity.