研究目的
To develop and test calibration models for integrated prediction of external and internal quality of 'Marsh' grapefruit using Vis/NIR spectroscopy.
研究成果
Vis/NIR calibration and validation results demonstrated that rind and internal quality parameters of 'Marsh' grapefruit could be predicted. It is therefore suggested that Vis/NIRS could be used as a non-destructive method to evaluate fruit quality in order to deliver fruit without rind disorder at market. Further investigation of quality parameters such as TPC and carbohydrate is suggested.
研究不足
The predictive value for parameters such as sucrose, glucose, fructose, TCP flavedo and TCP albedo were poor with r2 value of 0.12, 0.49 and 0.34, respectively. Reasons for these discrepancies are unknown and worth further investigations.
1：Experimental Design and Method Selection:
Visible to near infrared spectroscopy (Vis/NIRS) was used to acquire spectral information from 522 intact fruit. Principal component analysis (PCA) was applied to analyse spectral data to identify outliers. Savitzky-Golay second derivative with second order polynomial was employed as pre-processing method to correct light scattering properties of the spectra. Partial least square regression (PLSR) was used as the chemometric tool to develop models for predicting each quality attribute of intact 'Marsh' grapefruit.
2：Sample Selection and Data Sources:
'Marsh' grapefruit were harvested at commercial maturity from Bolton Citrus Farm, KwaZulu-Natal, and Unifruitti Farm, Mpumalanga provinces of South Africa. Fifteen individual trees were identified from each orchard for sampling. From each tree, twenty fruit of uniform size were randomly harvested and each fruit was used as an individual replicate.
3：List of Experimental Equipment and Materials:
FT-NIR spectrometer (MATRIXTM - F (Q410/A)), Minolta NR-4000 colorimeter, Labogene Scan Vac Cool Safe CS55-4 freeze-dryer, Milli-Q Integral Water Purification System.
4：Experimental Procedures and Operational Workflow:
Spectra were collected from two opposite sides along the equatorial region of each fruit after manually rotating the fruit 180° and the spectra were averaged. Reference quality parameters including colour indices were measured in (luminosity (L*), greenness (a*) and yellowness (b*)) colour space at three equidistant points around the equatorial axis of each fruit.
5：Data Analysis Methods:
Pre-processing, calibration and validation of Vis/NIRS spectral were performed using the Unscrambler chemometric software. The spectra were subjected to a test set validation by categorising the spectra into calibration (60%) set and validation (40%) set.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
