Services on Demand
Journal
Article
Indicators
-
Cited by SciELO
-
Access statistics
Related links
-
Cited by Google
Similars in SciELO
-
Similars in Google
Share
Chilean journal of agricultural research
On-line version ISSN 0718-5839
Chilean J. Agric. Res. vol.69 no.4 Chillán Dec. 2009
http://dx.doi.org/10.4067/S0718-58392009000400015
Chilean Journal of Agricultural Research 69(4):588-592 (October-December 2009)
SCIENTIFIC NOTE
Potential of Spectral Reflectance as Postharvest Classification Tool for Flower Development of Calla Lily (Zantedeschia aethiopica (L.) Spreng.)
Potencial de
Antonio J. Steidle Neto 1*, José A.S. Grossi 1, Daniela C. Lopes 2, and Enrique A. Anastácio 1
1 Universidade Federal de Viçosa, Departamento de Engenharia Agrícola, Secretaria de Pós-Graduação, Av. Peter Henry Rolfs s/nº, Campus Universitário, CEP: 36.570-000, Viçosa, Minas Gerais, Brasil. *Corresponding author ([email protected]).
2 Universidade Federal de São João del-Rei, Campus Alto Paraopeba, Rodovia MG
Received: 06 October 2008. Accepted: 02 December 2008.
ABSTRACT
Unsuitable postharvest management is one of the most serious problems that floriculture has to face. An option for reducing postharvest losses is to use automatic systems for flower sorting and classification, which yield consistent results, reduce costs and speed up these tasks. The objective of this work was show the potential of spectral reflectance to distinguish different postharvest development stages of calla lily flowers, Zantedeschia aethiopica (L.) Spreng., aiming the use of this technology within automatic systems for flower classification. The measuring equipment was a spectrometer connected to a portable computer and configured for reflectance data acquisition in the 400 to 1000 nm range. Based on the results, it was verified a differentiation between the spectral reflectance curves of calla lily flowers, with gradual decreases on the measured values according to the increase of the senescence stages. Thus, the spectral reflectance has potential to be used in the development of automatic systems for postharvest classification of calla lily flowers.
Key words: cut flowers, Zantedeschia aethiopica, spectrometer.
RESUMEN
El manejo poscosecha inadecuado es uno de los problemas más serios que la floricultura tiene que enfrentar. Una opción para reducir las pérdidas poscosecha es emplear sistemas automáticos para ordenar y clasificar las flores, los cuales permiten resultados consistentes, reducen gastos y aceleran estas tareas. El objetivo de este trabajo fue demostrar el potencial de la reflectancia espectral para discriminar las diferentes fases de desarrollo poscosecha de flores de cala (Zantedeschia aethiopica [L.] Spreng.), visando el uso de esta tecnología en sistemas automáticos para la clasificación de flores. El equipo de medición fue un espectrómetro conectado a un computador portátil y configurado para la adquisición de datos de reflectancia comprendidos en la región espectral de
Palabras clave: flores de corte, Zantedeschia aethiopica, espectrómetro.
INTRODUCTION
Cut flower agribusiness is a promising activity and has great development perspective considering internal and external markets. However, it is essential to improve the product quality and sanity in order to meet the international market demand (Mutersbaugh et al., 2005). Among the problems that floriculture has to face, the unsuitable postharvest management is one of the most serious. In hot and wet tropical countries, as
Most of flower classification job is performed by trained technicians, what could limit the improvement of services that fulfill the consumers' exigencies. Human visual classification is prone to mistakes since it is a subjective, monotonous, stressing and time-consuming task (Brosnan and Sun, 2002). An option for improving and speeding up the flower classification process is substitute the human operator by an automatic system based on the spectral reflectance of flowers. Among the characteristics of the study of the interaction between solar radiation and vegetation, spectral comportment of plants is highlighted, including the reflectance as the most analyzed phenomenon (
Based on spectral reflectance of vegetation is possible to identify differences and similarities between plants of same species (Vaiphasa et al., 2007), nutritional levels of plants (Min et al., 2008) and presence of pests and diseases (Goel et al., 2003). Information of reflectance can be obtained by satellites, images acquired from digital cameras coupled to aerial or terrestrial systems, spectrometers or other similar equipment (Gitelson et al., 2002).
In this context, the objective of this work was evaluating the potential of spectral reflectance of the spathes of calla lily flowers (Zantedeschia aethiopica (L.) Spreng.) to discriminate different postharvest development stages, aiming the use of this technology within automatic systems for flower classification.
MATERIALS AND METHODS
Ten calla lily flowers were harvest on same day of reflectance measurements. The flowers were sorted in crescent order according to their postharvest development stages, starting with opening flowers and finishing with the senescent ones (Figure 1).
Figure 1. Digital images of calla lily flowers sorted according different postharvest development stages.
During the trial these flowers were conditioned in a recipient with water. Each stem was cut as closely as possible to the floral button before beginning reflectance measurements. Thus, it was possible to completely extend the flower spathe along the measurement table. Spathes were divided into four quadrants and punctual measures of reflectance were made in each one of them (Figure 2).
Spectral reflectance measures of calla lily flowers were carried out at the Laboratory of Artificial Vision and Machine Projects, located at the Department of Agricultural Engineering of the Federal University of Viçosa, Brazil. It was used a spectrometer (model SD-2000, Ocean Optics Inc.,
Figure 2. Scheme of the equipment used for measuring the spectral reflectance of calla lily flowers.
The spectrometer probe was composed by a bunch with seven optical fibers. Six of them are responsible for the light emission, generated by a controlled source (model LS-1 Tungsten Halogen, Ocean Optics Inc.,
The reflectance values acquired by the spectrometer were calibrated by the software package OOIBase 32 (version 2.0), before be stored in computer. During the data acquisition, spectral reflectance was expressed as a relative percentage of the white pattern, as shown by equation described by Xing and Baerdemaeker (2005):
where ρλ is the spectral reflectance of the flower (%), Sλflower is the intensity of the reflected energy in the flower (dimensionless), Dλdark is the intensity of the reflected energy considering light absence (dimensionless), ρλwhite is the spectral reflectance of the Spectralon mini panel (dimensionless). All these parameters are applied considering a wavelength λ.
The intensity of the reflected energy considering light absence was obtained by obstructing the light input at the aluminum holder and turning off the light source. Also, the total reflectance of each spathe was calculated by using the arithmetic average of its four quadrants.
To analyze statistically the postharvest development stages of calla lily flowers, a randomized experimental design was used with 10 treatments (postharvest development stages) and four replications (spectral reflectance measures of the four quadrants of each spathe). For each treatment it was plotted a spectral reflectance curve based on the average values of the measures, considering wavelengths in the 400 and 1000 nm range. Using the SigmaPlot software (version 10.0) a nonlinear regression analysis based on the Hill sigmoidal model with three parameters was applied to each spectral reflectance curve. With this model, coefficients of determination (R2) greater than 95% were obtained for all regressions. Finally, the coefficients of the three parameters obtained in each regression were submitted to a confidence interval analysis (P < 0.05).
RESULTS AND DISCUSSION
In a general way, reflectance was high for all flowers, so much for spectral range of visible light (400 to 700 nm) as for spectral range of infrared radiation (700 to 1000 nm) (Figure 3). This result can be mainly justified by the presence of pigments, called flavonoids, which are responsible for the white color of the spathe of calla lily flowers. These pigments absorb radiation in the ultraviolet region of spectrum and reflect in the visible one (Yu et al., 2006). However, in spite of the similar spectral comportment of the curves, it was verified that the measured values decreased with the increase of senescence. Gates and Tantrapporn (1952), cited by Ponzoni (2001), also verified that the reflectance of leaves in the visible region of spectrum decreases with the senescence intensification.
Figure 3. Spectral reflectance curves of calla lily flowers.
Moreira (2007) affirmed that the main factors that influence spectral comportment of leaves are the amount and types of pigments, water and air contents, internal cellular structures and maturation and senescence stages. This affirmation is also valid for petals and sepals since, according to Esau (2000), these floral organs are similar to leaves when considering their internal structure and biochemical composition.
Specifically in the spectral region comprehended between 400 and 460 nm, it was observed an expressive increase of reflectance, which was more accentuated for flowers in initial development stages. From 460 to 920 nm, high and about constant values of spectral reflectance were verified. In the other hand, observing the near infrared region (920 to 1000 nm), it was verified progressive reductions in the reflectance of all calla lily flowers. This decreasing tendency can be attributed to the radiation absorbed by the water contained in spathes. According to Niemann et al. (2002), one of the absorption peaks occurs at 970 nm.
Flowers in advanced senescence presented an increase in reflectance at wavelengths between 800 and 920 nm (Figure 3, flowers 7 to 10). This rise can be explained by the changes in the optical properties of spathes due to reflection, transmission and absorption of incident radiation. That is, during the senescence process, the yellow pigmentation of the external borders of spathes was evidenced, as well as, according to Moreira (2007), an increase in the intercellular spaces of spathes occurred.
The confidence intervals of the three parameters obtained with the Hill sigmoidal model for each nonlinear regression are presented in Table 1. Despite some confidence intervals associated with parameters B and C overlap, the spectral reflectance curves were considered statistically different due to the confidence intervals of parameter A never overlap (P < 0.05).
Table 1. Confidence intervals of the three parameters of the Hill sigmoidal model for each nonlinear regression associated with the postharvest development stages of calla lily flowers.
Figure 4 presents the spectral reflectance curves of the calla lily flowers for typical wavelengths of blue (475 nm), green (510 nm), red (650 nm) and near infrared (850 nm). Despite wavelength band of the near infrared extends to around 2500 nm, the average value of the 700 to 1000 nm range was used because of a limitation in the measurement span of the spectrometer.
The spectral reflectance curve associated with the wavelength of blue presented the greater variation considering all postharvest development stages of calla lily flowers. However, the spectral reflectance values associated with the wavelength of near infrared allowed a best differentiation between the postharvest development stages from 1 to 5. Besides, for distinguishing the postharvest development stages from 6 to 10, it was verified that the spectral reflectance of the wavelength of blue presented best results (Figure 4).
Figure 4. Spectral reflectance curves of calla lily flowers for typical wavelengths of blue, green, red and near infrared.
CONCLUSIONS
Based on the results, it could be affirmed that the different postharvest development stages are characterized by temporal and intrinsic changes in calla lily flowers, which also impose significant changes in spectral reflectance. Therefore, it was concluded that spectral reflectance has large potential to distinguish different postharvest development stages and can be used within automatic systems for calla lilies classification.
ACKNOWLEDGEMENTS
The authors would like to thank the FAPEMIG (Fundação de Amparo à Pesquisa do Estado de Minas Gerais) for the financial support.
LITERATURE CITED
Brosnan, T., and D.W. Sun. 2002. Inspection and grading of agricultural and food products by computer vision systems: a review. Comput. Electron. Agr. 36: 193-213.
Dickinson, R.E. 2008. Determination of the multi-scattered solar radiation from a leaf canopy for use in climate models. J. Comput. Phys. 227: 3667-3677.
Esau, K. 2000. Anatomia das plantas com sementes. 293 p. Edgard Blücher, Editora da Universidade de São Paulo, São Paulo, Brasil.
Gitelson, A.A., R. Stark, U. Grits, D.C. Rundquist, Y.J. Kaufman, and D.
Goel, P.K., S.O. Prasher, J.A. Landry, R.M. Patel, R.B. Bonnell, A.A. Viau, and J.R. Miller. 2003. Potential of airborne hyperspectral remote sensing to detect nitrogen deficiency and weed infestation in corn. Comput. Electron. Agr. 38:99-124.
Min, M., W.S. Lee, T.F. Burks, J.D. Jordan, A.W. Schumann, J.K. Schueller, and H. Xie. 2008. Design of a hyperspectral nitrogen sensing system for orange leaves. Comput. Electron. Agr. 63:215-226.
Moreira, M.A. 2007. Fundamentos do sensoriamento remoto e metodologias de aplicação. 320 p. 3rd ed. Editora da Universidade Federal de Viçosa, Viçosa, Minas Gerais, Brasil.
Mutersbaugh, T., D. Klooster, M.C. Renard, and P. Taylor. 2005. Certifying rural spaces: quality-certified products and rural governance. J. Rural Stud. 21:381-388.
Niemann, K.O., D.G. Goodenough, and A.S. Bhogal. 2002. Remote sensing of relative moisture status in old growth Douglas fir. Int. J. Remote Sens. 23:395-400.
Ponzoni, F.J. 2001. Comportamento espectral da vegetação. p. 157-
Vaiphasa, C., A.K. Skidmore, W.F. De Boer, and T. Vaiphasa.
Weinberger, K., and T.A. Lumpkin. 2007. Diversification into horticulture and poverty reduction: a research agenda. World Dev. 35:1464-1480.
Xing, J., and J.D. Baerdemaeker. 2005. Bruise detection on 'Jonagold' apples using hyperspectral imaging. Postharvest Biol. Technol. 37:152-162.
Yu, O., M. Matsuno, and S. Subramanian. 2006. Flavonoid compounds in flowers: Genetics and biochemistry. 300 p. Global Science Books,