Tesis doctoral de Ana María Herrero Langreo
This work is centred in the use of spectral optical methods for the assessment of peach ripeness. It aims to contribute to fulfil an increasing need for controlling peach quality, a major complain between consumers and severely affected by the lack of ripeness. the measurement of optimum peach ripeness after harvest is a controversial issue, which needs to be balanced between a minimum ripeness, acceptable for the consumer, and a maximum ripeness, to minimize fruit losses during the postharvest process. Preferably, peach quality assessment methods should be quick, non destructive, and applicable on line. These traits would allow sampling big proportions of each population. This is especially important in peach fruit due to the high ripeness variability within populations from the same harvest and after the same postharvest treatments. spectral optical methods do fulfil these characteristics and are thus, considered as an appropriate technique for this task. In addition, they can be related to a reliable indicator of peach ripening, the degradation of chlorophyll (chlorophyll presents maximum absorption at 680 nm). Furthermore, spectral imaging measurements, as opposed to punctual measurements, can provide spatially detailed information about peach ripeness. This is important in peach fruit because certain regions of the fruit, as the tips and shoulders generally ripen before, increasing the susceptibility of the fruit to transport and handling damages. consequently, the main objective of this thesis is the evaluation of optical spectral techniques for peach ripeness and quality assessment. Local and imaging multi and hyperspectral measurements are considered, as well as their combination with non destructive mechanical techniques. This research is aimed to support the development of classification procedures for grading peach fruit into quality classes and thus, to contribute to decision supporting systems for peach postharvest process. the experimental work was carried out between june and september 2005-2008. Fruits from different harvest dates, before and after postharvest ripening process were analyzed. this work was performed in three stages: in a first stage a classification methodology was developed and implemented to classify multispectral images of reflectance values at 680 nm (r680) over reflectance at 800 nm (r800). The proposed procedure was able to classify r680/r800 images into four ripeness classes, according to a non supervised classification and further assignment of anonymous images onto the previously generated classes. This classification was related to magness-taylor firmness (mtf), which is the main reference for peach commercial handling. The system was able to segregate 4 ripeness classes. These classes were related with peach handling recommendations: 85% of the samples from the most unripe class in validation were over the recommended firmness for harvest (mtf >45 n) and 73% from the samples of the ripest class were too soft for fruit post-harvest processing (mtf <18 n). additionally, the classification system was used for monitoring peach ripeness evolution through postharvest. Peach ripeness before and after the ripening process were compared, being able to identify up to 65% of immature fruits which will not ripen during postharvest. second and third works, were carried out in collaboration with the research institution information and technologies for agro-processes (itap), cemagref, montpellier, france. in an initial approach, two additional optical indexes were proposed (ind1 e ind2) based in different spectral corrections of the chlorophyll band (680 nm). These indexes were acquired through hyperspectral imaging and evaluated on both their aptitude to measure ripeness and their susceptibility to the effect of convexity of the fruit in the index values of the image. The newly proposed indexes as well as others found in literature were compared, obtaining the best results for ind2. finally, analysis procedures and optical indexes used in the above works were integrated with non destructive mechanical measurements. These variables were combined through multivariate analysis in order to estimate maturity and firmness. Optical and mechanical measurements were observed to be complementary for these tasks. Magness-taylor firmness was estimated in validation with an r2 of 0.76 and a root square standard error of prediction of 5.55 n (10.9% of the observed range: 20.5 to 71.6 n). Considering mean index values of sequential harvest dates as references of maturity, optical measurements were able to register fruit maturity stage. Mechanical variables, in contrast, were affected both by postharvest transport incidences and by maturity, showing thus, their potential for peach postharvest control.
Datos académicos de la tesis doctoral «Técnicas de imagen espectral para la supervisión y predicción de la madurez de melocotón en recolección y postrecolección«
- Título de la tesis: Técnicas de imagen espectral para la supervisión y predicción de la madurez de melocotón en recolección y postrecolección
- Autor: Ana María Herrero Langreo
- Universidad: Politécnica de Madrid
- Fecha de lectura de la tesis: 25/06/2010
Dirección y tribunal
- Director de la tesis
- Belen Diezma Iglesias
- Tribunal
- Presidente del tribunal: margarita Ruiz altisent
- Ana María Garrido varo (vocal)
- Mª coral Ortiz sanchez (vocal)
- luigi Bodria (vocal)