The objective of this work was to evaluate the feasibility of using hyperspectral imaging to discriminate artichokes from 2 cultivars and different harvest times. The cultivar ‘Catanese di Brindisi’ produced in Puglia is recognized as PGI, and therefore the need for a fast tool of discrimination of artichoke cultivars, also in consideration of the high economical value of this crop which normally is not propagated by seeds. ‘Catanese di Brindisi’ artichokes were collected in 400 samples (n=308 for calibration set and n=92 for test set) while ‘Violetto Foggiano’ artichoke were collected 7 times, for a total of 320 samples (n=244 for calibration and n=76 for test set). Spectral information in the Vis-NIR region were acquired using a spectral scanner (DV Srl, ver 1.4., Italia) with a detector working in reflectance mode from 400 to 1000 nm. SIMCA and PLS-DA were compared for the classification purpose, combined with preprocessing techniques. The best separation among the artichoke cultivars was achieved performing a second derivative whereas a combination of smoothing and MSC provided best classification with respect to harvest time. Conclusively, PLS-DA performed better then SIMCA with the latter resulting in low specificity values. In case of the cultivar classification, the sensitivity and specificity were 100% for the external test set. On the other hand, in case of the harvest dates the values of these parameters varied form 33-100 and 86-100%, respectively, and only 92 and 82% of the samples were correctly classified for calibration and test sets, respectively. The results encourage further studies on the application of hyperspectral imaging for the characterization of the origin of production, and of quality and to support producers to individuate the optimal harvest time.

Use of hyperspectral imaging for the discrimination of artichoke by cultivar and harvest time

Amodio, M. L.;Colelli, G.
2020-01-01

Abstract

The objective of this work was to evaluate the feasibility of using hyperspectral imaging to discriminate artichokes from 2 cultivars and different harvest times. The cultivar ‘Catanese di Brindisi’ produced in Puglia is recognized as PGI, and therefore the need for a fast tool of discrimination of artichoke cultivars, also in consideration of the high economical value of this crop which normally is not propagated by seeds. ‘Catanese di Brindisi’ artichokes were collected in 400 samples (n=308 for calibration set and n=92 for test set) while ‘Violetto Foggiano’ artichoke were collected 7 times, for a total of 320 samples (n=244 for calibration and n=76 for test set). Spectral information in the Vis-NIR region were acquired using a spectral scanner (DV Srl, ver 1.4., Italia) with a detector working in reflectance mode from 400 to 1000 nm. SIMCA and PLS-DA were compared for the classification purpose, combined with preprocessing techniques. The best separation among the artichoke cultivars was achieved performing a second derivative whereas a combination of smoothing and MSC provided best classification with respect to harvest time. Conclusively, PLS-DA performed better then SIMCA with the latter resulting in low specificity values. In case of the cultivar classification, the sensitivity and specificity were 100% for the external test set. On the other hand, in case of the harvest dates the values of these parameters varied form 33-100 and 86-100%, respectively, and only 92 and 82% of the samples were correctly classified for calibration and test sets, respectively. The results encourage further studies on the application of hyperspectral imaging for the characterization of the origin of production, and of quality and to support producers to individuate the optimal harvest time.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11369/393454
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact