Aristotle University Of Thessaloniki (AUTH)
Εκτίμηση της ποιότητας λαχανικών με τη χρήση αναλυτικών, οπτικών και τηλεπισκοπικών μεθόδων
Abstract
dc:descriptionIn the present PhD thesis, the efficacy of several non destructive techniques in assessing qualitative components, degree of ripening and freshness status of vegetables, as well as in discriminating the commodities according to the above traits was investigated. In particular, non destructive measurements were recorded on tomato, pepper fruit and in potato tubers with optical sensors such as colormeter and chlorophyll fluorometer, as well as with remote sensing ones such as Digital single-lens reflex R-G-B camera, digital compact R-G-NIR or R- B-NIR cameras, spectroradiometer (400- 25000 nm) and Vis (400- 1100 nm) spectrograph, in order to investigate their possibility of implementation, the accuracy, consistency and reliability of the prediction models. Tomato fruit classification and postharvest ripening monitoring based on maturity stage of fruit at the time of harvest are necessary, in order to guarantee the highest possible quality and marketability of the final ripe product. The aim of this study was to investigate, if three innovative and convenient, in terms of rapidness, accuracy and expense, non-destructive protocols can be exploited in identifying the maturity stage of tomato fruit, while also estimating the pericarp’s pigments content, as well as further monitoring the postharvest ripening of tomato fruit during storage. These protocols included the measurement chlorophyll fluorescence (OJIP), the non-photochemical quenching of photosystem (NPQ) and the light harvesting complex (LHC). The efficiency of these methods was compared to that of the colorimeter, which is widely used for such purpose. According to the results, specific parameters of all the three protocols can classify the tomato fruit according to their maturity stage, as efficiently as the hue angle parameter of the color measurements. Additionally, certain parameters of these protocols were identified that can monitor the tomato fruit ripening during storage more reliably even compared to the colorimeter parameters, while also being strongly correlated with the tomato fruit pericarp pigments’ content. Among the three protocols, the OJIP is the most accurate one which also excels in generating data faster than the rest. Moreover, the ‘fix area’ parameter of the OJIP protocol retains the efficiency to discriminate fruit during 16 days of storage according to their initial ripening stage at harvest, even when all tomatoes turn red, in contrast to hue angle parameter, which was less efficient in distinguishing the different initial maturity stages, beyond the 8th day of storage. Models based on ‘fix area’ parameter better reflect the actual ripening process of fruit per maturity stage than the ones based on hue angle data. The optimal quality, the marketability and the shelf-life of pepper fruit often depend on the ripening stage at harvest. Moreover, it is necessary to estimate the process of ripening in various steps of the supply chain. The potential of four non- destructive techniques as rapid tools for discriminating the process of ripening was investigated. Chlorophyll fluorescence, Vis/NIR spectroscopy, two commercial digital imaging cameras- one R-G-B dslr and a modified R-G-NIR compact camera were assessed as alternatives to color measurements in classifying pepper fruit according to their ripening stage. Freshly harvested mature bell pepper fruit ‘cv. Denver’ were sorted in six distinct ripening stages (S1-S6), from green to full red (mature green 100% green-S1, green to brown-S2, brown-S3, orange-S4, red-S5 and full red-S6 stage), by visually assessing their color. From the results, it is concluded that the a*/b* color parameter that was determined using a Chroma meter was proven to be the most accurate descriptor in ripening stage classification, either at harvest or during storage, as confirmed by chemometrics, such as principal component (PCA), partial least square (PLS) and support vector machine (SVM) analyses and mean comparisons, as well. Therefore, the increase of a*/b* ratio can be used as an indicator to estimate the degree of fruit ripeness. The a*/b* ratio was strongly correlated with fruit spectral reflectance and chlorophyll fluorescence data, as well as withtwo imaging indices generated by digital R-G-B and R-G-NIR imaging technique that are reported for the first time on fruit in this study. The above results indicate that it is possible to estimate the a*/b* ratio and simultaneously the ripening stage of pepper fruits with high accuracy and consistency, by using individually several non-destructive techniques. However, freshness status of bell pepper fruit can be reliably assessed, irrespectively of the ripening stage at harvest, only by spectral reflectance data even by processing specific wavelengths that were identified after implementing the genetic algorithm. The most significant nutritional components of bell pepper fruit (dry matter, soluble solids, ascorbic acid, soluble phenols, chlorophylls and carotenoids, as well as the total antioxidant capacity) in each ripening stage and during storage were determined. Several non-destructive techniques (color, chlorophyll fluorescence, Vis/NIR spectroscopy, image processing using a conventional R-G-B digital single-lens reflex (DSLR) and a modified R-G-NIR compact camera) as rapid tools for assessing the nutritional quality of peppers was investigated. Mature bell pepper fruit were sorted in six distinct ripening stages (S1-S6), from green to full red and were stored at 10oC for 8 days. The nutritional composition in peppers was mainly affected by the ripening stage of bell fruit at harvest and only slightly by the postharvest storage period. All of the non destructive techniques accurately predicted fruit internal quality (r= 0,835- 0,942) with each models’ accuracy being related to the individually assessed components. Fresh harvested potato tubers from three commercial varieties (Spunta, Banba and Sylvana) were stored for 5 months at 5 oC in darkness. During storage, color and chlorophyll fluorescence were measured, digital images with three compact cameras (R-G-B, R-G-NIR and R-B-NIR) were captured and proper image analysis was performed, as well as spectral reflectance was acquired both on the external periderm of the tuber and in the inner flesh part, as well. Partial least square regression and discriminant analysis, combined with multilinear regression and feature selection techniques (forward iPLS, variance inflation factor and variable importance scores) were implemented, in order to identify the optimal method to estimate freshness of the potatoes and discriminate them according to the genotype. The partial least square regression analysis of spectra reflectance at the vis/NIR region of the electromagnetic spectrum (400- 1000 nm) that was captured at three distinct areas on the potato tubers, externally or internally, produced the most reliable model in assessing potato freshness irrespectively of the variety, with a cross- validated regression coefficient (Rcv) equal to 0,916 and 0,929, as determined by external or internal measurements, respectively. Feature selection algorithms not only indicated specific regions of interest that mostly affected the accuracy of the model, but further processing identified 460 nm and 656 nm as the most informative wavelengths to provide the most accurate multilinear model in terms of highest probability, strongest regression and lowest collinearity and root mean cross validation error. Similarly, the best discrimination of the three potato varieties, was achieved with the spectral reflectance measurement on the skin of the tubers (93% of the cross-validation sample sets), while R-G-NIR and R-B-NIR image components combined with selected chlorophyll fluorescence parameters (Fi, Fv and Phi_Eo) performed better than color measurement in both potato freshness assessment irrespectively of the variety tested or cultivar classification irrespectively of the storage period.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Kasampalis, Dimitrios
- Κασαμπαλής, Δημήτριος
Subjects
dc:subject × 16- Μη καταστρεπτικές μέθοδοι
- Λαχανικά
- Εωτερική ποιότητα
- Φθορισμός χλωροφύλλης
- VIS/ NIR φασματοσκοπία
- Non-destructive methods
- Vegetables
- Internal quality
- Chlorophyll fluorescence
- VIS/NIR spectroscopy
- Γεωπονία, Δασολογία και Αλιεία
- Γεωπονικές Επιστήμες και Κτηνιατρική
- Γεωπονία, γενικά
- Agriculture, Forestry and Fisheries
- Agricultural and Veterinary Sciences
- Agriculture, general
Rights
dc:rights- Statement dc:rights
-
- BY_NC_SA_4_0
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/49267