Time-resolved reflectance spectroscopy nondestructively reveals structural changes in ‘Pink Lady®’ apples during storage

Available online at www.sciencedirect.com Available online at www.sciencedirect.com

Procedia Food Science

Procedia – Food Science 00 (2011) 000–000

Procedia Food Science 1 (2011) 81 – 89

www.elsevier.com/locate/procedia

11th International Congress on Engineering and Food (ICEF11)

Time-resolved reflectance spectroscopy nondestructively reveals structural changes in ‘Pink Lady®’ apples during storage Maristella Vanoli a,b*, Anna Rizzoloa, Maurizio Grassia, Andrea Farinab, Antonio Pifferib, Lorenzo Spinellic, Alessandro Torricellib a

The Agricultural Research Council – Food Technology Research Unit (CRA-IAA),via Venezian 26, I-20133 Milan, Italy b Politecnico di Milano, Dipartimento di Fisica, Piazza Leonardo da Vinci 31, I-20133 Milan, Italy c Istituto di Fotonica e Nanotecnologie – CNR,, Piazza Leonardo da Vinci 31, I-20133 Milan, Italy

Abstract With the aim of studying the optical properties of ‘Pink Lady®’ apples measured by TRS during storage in normal atmosphere and of relating them to fruit maturity and structural characteristics, 60 apples, measured at harvest by TRS at 670 nm and in the spectral range 740−1100 nm, were ranked on the basis of decreasing µa670 (increasing maturity) and randomized into 6 batches corresponding to 6 times of analysis: at harvest and after 7, 15, 29, 66 and 91 days of storage at 1°C. At each storage time, apples were measured by TRS in the same spectral range as at harvest, and firmness, intercellular space volume (RISV), soluble solid content (SSC) and starch hydrolysis were measured; Streif Index and weight loss were computed. Overall, the µa spectra showed a maximum at 670 nm (chlorophyll-a) and at 980 nm (water), while the µ’s spectra had a decreasing trend with the wavelength increase. The µa670 and µa980 decreased with storage time as a consequence of a decrease in chlorophyll and water contents, respectively. The size and the density of the scattering centers were affected by the interplay of various phenomena: starch hydrolysis, flesh softening, water loss and RISV increase. High correlation coefficients were found between µa in the spectral region 920−1100 nm and firmness, Streif Index and RISV, while low correlation coefficients were observed between µa and SSC, and between µ’s and all the quality parameters. Using PLS analysis, good prediction for firmness, RISV and Streif Index were obtained using µa spectra; models’ performance improved when µa and µ’s are combined. © Published by Elsevier B.V. Selection peer-review underpeer-review responsibility of 11th International Congress ©2011 2011 Published by Elsevier Ltd.and/or Selection and/or under responsibility of ICEF11 on Engineering and Food (ICEF 11) Executive Committee. Executive Committee Members. Keywords: apple structure, TRS, absorption coefficient, scattering coefficient, fruit maturity

* Corresponding author. Tel.: +39-02-239557210; fax: +39-02-2365377. E-mail address: [email protected].

2211–601X © 2011 Published by Elsevier B.V. Selection and/or peer-review under responsibility of 11th International Congress on Engineering and Food (ICEF 11) Executive Committee. doi:10.1016/j.profoo.2011.09.014

82

2

Maristella Vanoli/ Procedia et al. / Procedia Food Science 1 (2011) 81 – 89 Author name – Food Science 00 (2011) 000–000

1. Introduction Apple fruit is mainly composed of the fleshy tissue of parenchyma cells permeated with vascular tissue and intercellular air spaces enclosed by the epidermis. Each tissue has a different microstructure composition which determines the fruit mechanical properties affecting texture and gas (O2 and CO2) diffusivity determinant of fruit respiration. Cell size and shape, presence of an adhesive middle lamella between individual cells, turgor pressure, mechanical properties of the cell wall, the amount of intercellular spaces and subcellular features (plasma membrane and plasmodesmata) determine the macroscopic properties of fruit. The knowledge of all microstructural characteristic is important for the understanding of fruit quality at harvest and after storage, as well as the susceptibility to physiological disorders. The shape and size of these components show considerable variability according to cultivar, fruit development, harvest date and storage conditions [1-3]. A fruit can be modelled as a diffusive media, where light distribution is determined by the interplay between scattering phenomena (due to fruit microstructure) and light absorption (due to the presence of chromophores and other chemical compounds). Time-resolved reflectance spectroscopy (TRS) provides a complete optical characterization of diffusive media with the simultaneous non-invasive measurement of the optical properties of absorption and scattering. TRS is based on the measurement of the temporal delay and the broadening experienced by a short laser pulse (pulse duration in the order of 100 ps) while travelling through a turbid medium [4]. By using an appropriate theoretical model of light penetration for the analysis of photon time distribution, it is possible to simultaneously estimate the absorption coefficient (µa) and the reduced scattering coefficient (µ’s). Light penetration achieved by TRS in most fruit and vegetables can be as great as 1-2 cm, depending on the optical properties [5]. Hence, TRS provides information on the internal properties of the medium and is not significantly affected by surface features [6]. Previous studies on apples have shown that µa measured at 630 nm (µa630) is related to fruit maturity: more mature apples (low µa630 values) had lower titratable acidity at harvest and higher soluble solids after storage compared to less mature fruit (high µa630 values) and were perceived sweeter, more aromatic and were more appreciated by panellist at sensory analysis [7]. On the other hand, the reduced scattering coefficient gave an insight into the textural properties of apple fruit: µ’s measured at 750 and 780 nm were linked to pectin composition showing a high and positive correlation with galacturonic acid content in water soluble pectin fraction, and a negative correlation with residue insoluble pectin and protopectin index [8]. The µ’s measured in the range between 750 and 790 nm were also related to mechanical properties of fruit (firmness, stiffness, intercellular spaces) and to sensory attributed related to structure (firm, crispy, mealy and juicy) [9, 10]. The aim of this work was to study the optical properties of ‘Pink Lady®’ apples measured by TRS in the near infrared spectral range during storage in normal atmosphere, and to relate these optical properties to fruit maturity and structural characteristics. 2. Materials and Methods Sixty apples, picked in Laimburg (Bolzano, Italy), at harvest were individually weighed and measured by TRS at 670 nm and in the spectral range 740−1100 nm on two opposite positions around the equator region. Apples were ranked on the basis of decreasing µa670, averaged on the two fruit sides (increasing maturity), and randomized into 6 batches corresponding to 6 times of analysis: at harvest (0) and after 7, 15, 29, 66 and 91 days of storage at 1°C in normal atmosphere. At each storage time, apples were measured by TRS in the same spectral range as at harvest. In parallel, flesh firmness (11 mm diameter plunger mounted on an Instron UTM, crosshead speed 200 mm/min), intercellular space volume (RISV, [11]), soluble solid content (SSC) and starch hydrolysis (EUROFRU scale from 1=minimum to

83 3

MaristellaAuthor Vanoliname et al./ /Procedia Procedia– Food Science 00 1 (2011) – 89 Food Science (2011) 81 000–000

10=maximum starch hydrolysis, [12]) were measured. Streif Index (firmness/(SSC × starch hydrolysis) [13]) and weight loss were calculated. The sample corresponding to 66 days of storage was not considered due to a failure in the TRS system. TRS measurements were performed in the spectral range 670−1100 nm with a broadband setup employing a white light laser (SC450, Fianium, UK) for generation of light pulses (10ps duration, 40MHz repetition rate, 1mW/nm average power), a watercooled double microchannel plate photomultiplier (R1564U, Hamamatsu, Japan), and a time correlated single-photon counting board (SPC-130, Becker & Hickl, Berlin, Germany). The temporal resolution of the overall system, calculated as the FWHM of the instrumental response function, is less than 90 ps. Details on the setup can be found in D’Andrea et al. [14]. TRS data were analyzed by using a spectrally constrained approach [15]. The spectral dependence of the optical properties is inserted in the analytical solution for light transport in diffusive media and best fit is obtained by simultaneously considering TRS data at multiple wavelengths and using structure parameters (e.g. density and size of scatterers) and chromophore concentrations (e.g. chlorophyll and water) as free parameters. The relationship between absorption coefficient and tissue constituents is given by the Beer Law,

µ a (λ ) = ∑ ciε i (λ ) = cCHLε CHL (λ ) + cH 2Oε H 2O (λ ) + bkg

(1)

i

where cCHL and cH2O are the chlorophyll and water concentration, respectively, and bkg is a constant value to account for the contribution of other absorbers. An approximation to the Mie theory is used to relate the reduced scattering coefficient to the structural properties of the measured diffusive sample:

µ s′ (λ ) = a (λ / λ0 ) −b

(2)

where a is the scattering coefficient at wavelength λ0 (in our case we choose λ0 = 600 nm) and b is a parameter related to the size of scatterers. Quality parameter data were processed by analysis of variance and means were compared by Tukey’s test. Correlations between optical properties at the measured wavelengths and quality parameters were studied by using PROC CORR procedure (SAS/STAT, SAS Institute Inc., Cary, NC, 1999). Partial Least Squares regression analysis (PLS, The Unscrambler X, v..10.0.1, Camo, Norway) was developed to model spectra data and quality characteristics. Each model was validated by cross validation (16 segment validation with a random assignment of samples to the segment). 3. Results and Discussion 3.1. Fruit quality characteristics Considering the variation in the quality parameters during storage (Table 1), apple softening began after 29 days of storage when there was a significant increase of RISV. Starch hydrolysis was already in act after 7 days’ storage and the phenomenon continued up to 91 days at 1°C, when starch disappeared. At the same time, Streif Index gradually decreased and weight loss increased. Only SSC did not change with storage time.

84

Maristella Vanoli/ Procedia et al. / Procedia Food Science 1 (2011) 81 – 89 Author name – Food Science 00 (2011) 000–000

4

Table 1. Quality characteristics of ‘Pink Lady®’ apples during storage Firmness

RISV

Starch index

Streif index -1

SSC

Weight loss

days at 1°C

(N)

(%)

(N x °Bx )

(°Bx)

(%)

0

84.5 a

25.3 c

8.6 c

0.073 a

13.9 a

0.95 d

7

81.8 ab

24.8 c

9.3 b

0.067 ab

13.5 a

1.22 dc

15

84.2 a

24.9 c

9.5 ab

0.066 bc

13.8 a

1.53 c

29

77.1 b

27.7 b

9.9 a

0.059 c

13.5 a

2.38 b

91

52.3 c

31.1 a

10.0 a

0.038 a

14.1 a

4.69 a

3.2. Spectra characteristics Overall, the absorption spectra showed a maximum at 670 nm (chlorophyll-a) and at 980 nm (water), while the scattering coefficient slowly decreased with increasing wavelength (Figure 1). At harvest, the absorption coefficient at 670 nm ranged from 0.035 to 0.076 cm-1, showing very low values compared with other apple cultivars already studied [5, 7], but confirming the findings of Rizzolo et al. [14]; the µ’s values were similar to those previously found in other apple cultivars, ranging from 11.59 to 13.72 cm-1 [5, 10] (Figure 1). Comparing TRS spectra measured at harvest with those measured on the same fruit during storage, µa670 values were lower in apples stored for 15, 29 and 91 days than in the same fruit measured at harvest, while differences in µa980 values were evident only for 91 days stored fruit (Figure 1). As for the scattering behavior, the parameter a reported in Eq. 2 is proportional to the density of the scattering centers and determines the absolute value of µ’s, while the parameter b depends on the size of the scatterers showing values about 1/3 lower than those of a and determines the slope of the scattering spectra. Apples stored for 15, 29 and 91 days showed lower values of a and b than those of the same apples at harvest (Figure 1). Comparing the TRS spectra at different storage times, on average, µa670 and µa 980 significantly decreased in apples stored from 29 days up to the end of the storage period, showing the lowest values in apples stored for 91 days (Figure 1). The decrease of µa670 and µa980 values was due to a decrease in the chlorophyll and water contents, respectively (Table 2), the first related to fruit ripening and the latter to water loss. In fact, during cold storage, apples ripened, as they lost firmness and hydrolyzed starch, as shown above, and the ripening process is accompanied by a progressive decrease in chlorophyll content [7, 9]. In agreement with the very low µa670 values found in ‘Pink Lady®’ apples, the estimated chlorophyll content was also lower than that found by Cubeddu et al. [17]: chlorophyll in ‘Pink Lady®’ apples was about 1/2-1/3 lower than that of ‘Golden Delicious’ and ‘Granny Smith’ apples, respectively. Considering the water content, it seems that TRS underestimates water content; in fact at harvest dry matter in apples was 15.5 % (data not shown) and water content measured by TRS was 82.3%. During storage, maximum weight loss was 4.7 % at 91 days of storage; this means that apples lost about 6% of water, in contrast to data of TRS water content, from which the computed water loss was about 9%. Scattering values, on average, decreased after 29 and 91 days of storage (Figure 1), confirming the findings of Vanoli et al. [10] on ‘Jonagored’ apples stored for 5 months. At 29 days at 1°C, the size and the density of the scattering centers showed the lowest values, having a slight increase up to 91 days (Table 2). The changes in scattering parameters reflected the changes in the pulp structure during storage. The decrease of the parameter a up to 29 days of storage could be linked to apple softening; this process is accompanied by the enzymatic cell wall breakdown and so the density of the scattering particles in the fruit flesh decreased, as found by Qin and Lu [18] and Bobelyn et al. [19]. As a consequence of this

85 5

MaristellaAuthor Vanoliname et al./ /Procedia Procedia– Food Science 00 1 (2011) – 89 Food Science (2011) 81 000–000

phenomenon, the light scattering at the cell interfaces is reduced leading to less scattering events in the tissue [19]. At the same time, RISV increased and some water was lost due to fruit transpiration, so the cells in the pulp tissue became smaller with more air filled pores. The values of the parameter b decreased but scattering values increased, as there was an higher refractive index mismatch leading to more and stronger scattering events [1, 19]. Actually the situation in the pulp tissue is much more complex than just described, as the scattering centers of a fruit are not expected to be homogeneous spheres. These parameters do not assess the real size of scattering centers in the tissue, rather they are average equivalent parameters that could be related to physical characteristic of fruit. 17

7 days

0.3

µ's (cm‐1)

0.04

µa (cm ‐1)

0.5

670

0.2 0.0

800

600

µ's (cm‐1)

670

0.0 800

0.5

15 days

15 13 11

1000

600

µ's (cm ‐1)

670

0.2

800

17

29 days 0.04

µa (cm ‐1)

1000

9 600

0.0

1000

29 days

15 13 11 9

600

800

0.5

1000

600

17

670

0.2 0.0

µ's (cm ‐1)

91 days 0.04

µa (cm ‐1)

800

17

15 days

0.2

0.3

11

1000

0.04

µa (cm ‐1)

0.5

0.3

13

9 600

0.3

7 days

15

800

1000

91 days

15 13 11 9

600

800 1000 wavelenght (nm)

600

800 1000 wavelenght (nm)

Fig. 1. Absorption (left) and scattering (right) spectra of ‘Pink Lady®’ apples at harvest (solid line) and after storage (dotted line)

Maristella Vanoli/ Procedia et al. / Procedia Food Science 1 (2011) 81 – 89 Author name – Food Science 00 (2011) 000–000

6

Table 2. Chlorophyll content, water content and scattering parameters of ‘Pink Lady®’ apples during storage absorption

days at 1°C

scattering parameters

chlorophyll content

water content

a

b

(µM-1)

(%)

(cm-1)

(-)

0

0.435 a

82.3 ab

13.91 a

0.280 ab

7

0.480 a

83.0 ab

13.84 a

0.325 ab

15

0.400 ab

83.7 a

13.61 a

0.185 c

29

0.415 ab

82.1 b

12.91 a

0.178 c

91

0.287 b

73.7 c

13.13 a

0.245 b

By studying the correlations between optical properties and quality parameters (firmness, RISV, SSC, Streif Index), high correlation coefficients (r ≥ 0.8) were found between µa in the spectral region 9201100 nm and firmness (positively), Streif Index (positively) and RISV (negatively) (Figure 2), while low correlation coefficients (r ≤ 0.4) were observed between µa and SSC, and between µ’s and all the quality parameters (data not shown). Vanoli et al. [20] found high and positive correlations between µa 912 and sensory firmness, juiciness, crispness and percent juice and negative correlations with mealiness and RISV, while lower or no correlations were found between µa measured in the wavelength range 630-780 nm and quality parameters of apple fruit [8, 9]. On the contrary, the same authors found that scattering coefficients measured at 750 and 780 nm were well correlated to fruit texture both from the mechanical and sensory point of view and also to pectic composition. Probably in this work, µ’s decreased due to fruit softening, but, at the same time, it increased due to water loss. This scenario complicated the relationship between the measured spectra and the firmness values of stored apples, leading to a positive or negative correlation between optical properties and structural characteristics depending on which effect dominates the spectrum. 1.0 correlation coefficient ( r )

86

firmness

0.8

RISV

0.6

SSC

0.4

Streif I.

0.2 0.0 ‐0.2 ‐0.4 ‐0.6 ‐0.8 ‐1.0 650

750

850

950

1050

1150

wavelength (nm)

Fig. 2. Correlation coefficients between µa and quality characteristics of ‘Pink Lady®’ apples during storage

By using PLS to model optical properties and quality parameters, good prediction of firmness, RISV and Streif Index were obtained using absorption spectra, while bad results were obtained using scattering spectra (Table 4). When scattering was added to absorption the performance of the models improved, as

87 7

MaristellaAuthor Vanoliname et al./ /Procedia Procedia– Food Science 00 1 (2011) – 89 Food Science (2011) 81 000–000

R2 increased and RMSE decreased (Table 3). SSC was not satisfactory predicted by either absorption or scattering, as R2 of the model was not higher than 0.13 (data not shown). Table 3. Calibration performances for quality parameters of ‘Pink Lady®’ apples during storage Quality

TRS

Latents

Calibration

parameters

Parameter

Variables

R2cal

RMSEC

R2cv

RMSECV

Firmness (N)

µa

1

0.83

5.36

0.82

5.56

µ’s

4

0.18

11.72

0.02

13.02

µa + µ’s

4

0.88

4.50

0.83

5.44

µa

1

0.79

1.26

0.77

1.30

µ’s

3

0.27

2.34

0.14

2.55

µa + µ’s

4

0.87

1.00

0.82

1.19

µa

1

0.70

0.0071

0.70

0.0073

µ’s

4

0.15

0.0121

0.04

0.0130

µa + µ’s

4

0.81

0.0057

0.75

0.0068

RISV (%)

Streif Index

Cross-validation

The results of PLS also confirm the lack of correlations between scattering and quality parameters, even if Lu [21] found that both µa and µ’s measured by hyperspectral imaging technique on 600 stored apples were well correlated to fruit firmness with r = 0.83 and 0.70, respectively. It is interesting to note that when µa and µ’s are combined, better predictions of fruit firmness were achieved, as previously found by Rizzolo et al. [9], Valero et al. [22] and Lu [21]. 4. Conclusion Our results indicate that the optical properties of ‘Pink Lady®’ apples measured by TRS dramatically changed already after 1 month of cold storage reflecting the changes in chemical composition and in textural characteristics occurring in the fruit flesh. Absorption spectra showed the decrease in chlorophyll (linked to fruit maturity) and in water content (related to water loss). As expected, the absorption coefficients were well correlated to Streif Index, i.e. to fruit maturity, and to fruit texture (firmness and RISV). Despite the lack of correlation between scattering properties and fruit quality parameters, by adding scattering coefficients to the absorption ones, the performance of firmness, RISV and Streif index models built by PLS analysis were improved. Further researches are needed to better clarify the relationship between scattering spectra and textural characteristics of apple tissue at macro and microscopic levels in a wider range of apple cultivars with different maturity. Acknowledgements We wish to thank Dr. Angelo Zanella from the Agricultural Research Centre of Laimburg, Auer (Post), Italy for fruit management. This work was partially supported by FP7 project n. 226783 InsideFood “Integrated sensing and imaging devices for designing, monitoring and controlling microstructure of foods”.

88

Maristella Vanoli/ Procedia et al. / Procedia Food Science 1 (2011) 81 – 89 Author name – Food Science 00 (2011) 000–000

8

References [1] Schotsmans W, Verlinden BE, Lammertyn J, Nicolaï BM. The relationship between gas transport properties and the histology of apple. Journal of the Science of Food and Agriculture, 2004, 84: 1131−1140. [2] Mendoza F, Verboven P, Mebatsion HK, Kerclhofs G, Wevers M, Nicolaï BM. Three-dimensional pore space quantification of apple tissue using X-ray computed microtomography. Planta, 2007, 226: 559−570. [3] Mebastion KH, Verboven P, Ho QT, Verlinden BE, Nicolaï BM. Modelling fruit (micro) structures, why and how? Trends in Food Science & Technology, 2008, 19: 59−66. [4] Torricelli A, Spinelli L, Contini D, Vanoli M, Rizzolo A, Eccher Zerbini P. Time-resolved reflectance spectroscopy for nondestructive assessment of food quality. Sensing and Instrumentation for Food Quality and Safety, 2008, 2: 82−89. [5] Cubeddu R, D’Andrea C, Pifferi A, Taroni P, Torricelli A, Valentini G, Ruiz-Altisent M, Valero C, Ortiz C, Dover C, Johnson D. Time-resolved reflectance spectroscopy applied to the non-destructive monitoring of the internal optical properties in apples. Applied Spectroscopy, 2001, 55(10): 1368−1374. [6] Saeys W, Velazco-Roa MA, Thennadil SN, Ramon H, Nicolaï BM. Optical properties of apple skin and flesh in the wavelength range from 350 and 2200 nm. Applied Optics, 2008, 47: 908−919. [7] Vanoli M, Eccher Zerbini P, Grassi M, Rizzolo A, Fibiani M, Pifferi A, Spinelli L, Torricelli A, Cubeddu R. The quality and storability of apples cv ‘Jonagored’ selected at harvest by time-resolved reflectance spectroscopy. Acta Horticulturae, 2005, 682(2): 1481−1488. [8] Vanoli M, Eccher Zerbini P, Spinelli L, Torricelli A, Rizzolo A. Polyuronide content and correlation to optical properties measured by time-resolved reflectance spectroscopy in ‘Jonagored’ apples stored in normal and controlled atmosphere. Food Chemistry, 2009, 115: 1450−1457. [9] Rizzolo A, Vanoli M, Spinelli L, Torricelli A. Sensory characteristics, quality and optical properties measured by timeresolved reflectance spectroscopy in stored apples. Postharvest Biology and Technology, 2010, 58: 1−12. [10] Vanoli M, Rizzolo A, Grassi M, Zanella A, Torricelli A, Spinelli L, Eccher Zerbini P. Relationship between scattering properties as measured by Time-resolved Reflectance Spectroscopy and quality in apple fruit. 3rd CIGR Section VI International Symposium on Food and Agricultural Products: Processing and Innovations. Naples (Italy), 24-26 September 2007. CD-ROM Proceedings (13 pages). [11] Baumann H, Henze J. Intercellular space volume of fruit. Acta Horticulturae, 1983, 138: 109−111. [12] Planton G. Le Point n.6, CTIFL, Paris, August 1995. [13] Streif J. Optimum harvest date for different apple cultivars in the ‘Bodensee’ area.. In: de Jager A., Johnson D. Höhn E. (Eds). The Postharvest Treatment of Fruit and Vegetables: Determination and Prediction of Optimum Harvest Date of Apples and Pears. COST 94, Brussels, Belgium, 1996, pp. 15−20. [14] D’Andrea C, Nevin A, Farina A, Bassi A, Cubeddu R. Assessment of variations in moisture content of wood using timeresolved diffuse optical spectroscopy. Applied Optics, 2009, 48(4): B87−B93. [15] D'Andrea C, Spinelli L, Bassi A, Giusto A, Contini D, Swartling J, Torricelli A,. Cubeddu R. Time-resolved spectrally constrained method for the quantification of chromophore concentrations and scattering parameters in diffusing media. Optics Express, 2006, 14 (5): 1888−1898. [16] Rizzolo A, Vanoli M, Cortellino G, Spinelli L, Torricelli A. Quality characteristics of air-dried apple rings: influence of storage time and fruit maturity measured by time-resolved reflectance spectroscopy. Proceedings 11th International Congress on Engineering and Food “Food Process Engineering in a Changing World” PS Taoukis, NG Stoforos, VT Karathanos, GD Saravacos (Eds.) NTUA, School of Chemical Engineering, Athens (GR), 2011, Vol 2, pp. 987−988. [17] Cubeddu R, D’Andrea C, Pifferi A, Taroni P, Torricelli A, Valentini G, Dover C, Johnson D, Ruiz-Altisent M, Valero C. Nondestructive quantification of chemical and physical properties of fruits by time-resolved reflectance spectroscopy in the wavelength range 650-1000 nm. Applied Optics, 2001, 40(4): 538−543. [18] Qin J, Lu R. Measurement of the optical properties of fruits and vegetables using spatially resolved hyperspectral diffuse reflectance imaging technique. Postharvest Biology and Technology, 2008, 49: 355−365. [19] Bobelyn E, Serban AS, Nicu M, Lammertyn J, Nicolaï BM, Saeys W. Postharvest quality of apple predicted by NIRspectroscopy: study on the effect of biological variability on spectra and model performance. Postharvest Biology and Technology, 2010, 55: 133−143.

Maristella Vanoliname et al./ Procedia / Procedia FoodScience Science00 1 (2011) 81 – 89 Author – Food (2011) 000–000 [20] Vanoli M, Rizzolo A, Eccher Zerbini P, Spinelli L, Torricelli A. Non-destructive detection of internal defects in apple fruit by Time-resolved Reflectance Spectroscopy. In: Nunes C (Ed.) “Enviromentally Friendly and Safe Technologies for Quality of Fruits and Vegetables”, Universidade do Algarve, Faro, Portugal, 2010, pp. 22−28. [21] Lu R. Spectroscopic technique for measuring the texture of horticultural products:spatially resolved approach. In: Zude M. (Ed). Optical monitoring of fresh and processed agricultural crops. CRC Press, Taylor & Francis Group, 2009, .pp 391−423. [22] Valero C, Ruiz-Altisent M, Cubeddu R, Pifferi A, Taroni P, Torricelli A, Valentini G, Johnson DS, Dover C. Selection models for the internal quality of fruit based on time domain laser reflectance spectroscopy. Biosystems Engineering, 2004, 88: 313−323.

Presented at ICEF11 (May 22-26, 2011 – Athens, Greece) as paper FMS105.

89 9