Non -destructive measurement of citrus internal quality using near infrared spectroscopy and x-ray computed temography

Lebotsa, Phalane Sekina (2017-03)

Thesis (MScEng)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: Quality evaluation of fruit remains an important aspect in the fruit industry. It is for this reason that non-destructive methods for quality evaluation of citrus fruit were investigated. Literature reviews revealed the possibilities of near infrared (NIR) spectroscopy to detect physicochemical properties and X-ray computed tomography (X-ray CT) to investigate internal structure of citrus fruit. The non-destructive methods were also used to classify the citrus fruit based on the quality parameters. First, Fourier transform near infrared (FT-NIR) spectroscopy was used to detect the physicochemical properties, mass, titratable acidity (TA), total soluble solids (TSS) and hue angle (h*) of ‘Satsuma’ mandarin. Partial least squares (PLS) analysis gave accurate results for mass, TSS and h*. The analysis showed that exclusion of spectra collected from stylar-end and stem-end resulted in improved prediction models. The exclusion of stylar-end and stem-end spectra, is therefore, an advantage as the two parts of the fruit may not contain much information about the fruit and therefore including them compromised quality of the information. When models developed from spectra of intact fruit were compared to models developed from peeled fruit, the removal of peel did not give an improvement which was attributed to heterogeneity of a peeled fruit surface and its transparency. OPLS-DA models were developed to distinguish between low and high classes of ‘Satsuma’ mandarin fruit, yielding accurate classification based on mass, TSS and h*. Secondly, granulation in ‘Navel’ orange and seeds in ‘Nadorcott’ madarin using X-ray CT based on radiographic images and tomographic images were investigated. Granulation was observed in ‘Navel’ orange radiographic and tomographic images. Accurate classification based on both image datasets was achieved. However, radiograhic images could not reveal seeds in ‘Nadorcott’ mandarins. The tomographic images, in contrary, were able to reveal ‘Nadorcott’ seeds despite its inability to clearly separate the seeds from the endocarp of the fruit. The study demonstrated the potential of NIR spectroscopy and X-ray CT as possible non-destructive technologies for sorting of fruit based on physicochemical properties and internal structures respectively.

AFRIKAANSE OPSOMMING: Geen opsomming beskikbaar

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/101203
This item appears in the following collections: