DIELECTRIC PROPERTIES OF HONEYDEW MELONS
AND CORRELATION WITH QUALITY
Wen-chuan Guo1 , Stuart O. Nelson2, Samir Trabelsi2, and Stanley J. Kays3
Northwest Agriculture and Forestry University,
College of Mechanical and Electronic Engineering, Yangling, China 712100
2
U. S. Department of Agriculture, Agricultural Research Service,
Russell Research Center, P. O. Box 5677, Athens, Georgia 30604-5677, USA
3
The University of Georgia,
Department of Horticulture, Athens, Georgia 30602, USA
1
Dielectric properties of three honeydew melon cultivars, grown and harvested to provide a
range of maturities, were measured with an open-ended coaxial-line probe and impedance analyzer
over the frequency range from 10 MHz to 1.8 GHz. Probe measurements were made on the external
surface of the melons and also on tissue samples from the edible internal tissue. Moisture content
and soluble solids content (SSC) were measured for internal tissue samples, and SSC (sweetness)
was used as the quality factor for correlation with the dielectric properties. Individual dielectric
constant and loss factor correlations with SSC were low, but a high correlation was obtained
between the SSC and permittivity from a complex-plane plot of dielectric constant and loss factor,
each divided by SSC, for both the external surface and internal tissue measurements. However,
SSC prediction from the dielectric properties by these relationships was not as high as expected.
Permittivity data (dielectric constant and loss factor) for the melons are presented graphically to
show their relationships with frequency for external surface and internal tissue measurements. A
dielectric relaxation for the external surface measurements, which is attributable to bound water and
Maxwell-Wagner relaxations, is also illustrated. Coefficients of determination for complex-plane
plots, moisture content and SSC relationship, and penetration depth are also shown graphically.
Further studies are needed for determining the practicality of sensing melon quality from the
dielectric properties.
Submission Date: 10 October 2006
Acceptance Date: 19 June 2007
Publication Date: 15 August 2007
INTRODUCTION
Rapid and reliable nondestructive
determination of quality and related
characteristics of agricultural products can be
helpful to producers, handlers and processors,
those marketing the produce, and consumers.
Visible and physical characteristics of many
Keywords: Dielectric properties, melons, quality, soluble
solids, permittivity, dielectric relaxation
41-2-44
fresh fruits and vegetables are available for
correlation with quality, and some of these,
such as color, size, weight, density, elasticity,
and firmness are used in automatic sorting of
some produce into different categories for the
market [Guo et al., 2001], [Kays and Paull,
2004]. For the honeydew melon, however, no
useful characteristics of this type have been
found for reliable correlation with quality.
Electrical properties of fruits and
vegetables known as dielectric properties have
Journal of Microwave Power & Electromagnetic Energy
Vol. 41, No. 2, 2007
been investigated in several studies. Seaman
and Seals [1991] found that the values of
microwave dielectric properties were different
for the surface and internal tissues of fruits.
Microwave dielectric properties of fresh
fruits and vegetables have been considered
for potential use in nondestructive sensing of
quality factors, such as maturity in peaches
and chilling injury in sweetpotatoes [Nelson,
1980]. Possible permittivity-based maturity
indices were identified for tree-ripened peaches
[Nelson et al., 1995], but much further research
was needed to assess the practicality of such
techniques.
Other measurements provided dielectric
properties data for fresh fruits and vegetables
over the frequency range from 200 MHz to 20
GHz [Nelson et al., 1994] and over the range
from 10 MHz to 1.8 GHz [Nelson, 2003]. Data
for tree-ripened peaches over the 200-MHz to
20-GHz range indicated an interest in such data
at frequencies below 200 MHz [Nelson et al.,
1995]. Therefore, the frequency range of 10
MHz to 1.8 GHz was selected for dielectric
spectroscopy measurements of honeydew
melon tissue to learn whether differences in
the dielectric properties might be correlated
with quality [Nelson et al., 2006]. These
same data were studied in relation to dielectric
relaxation in complex systems and subjected
to advanced mathematical analysis and curve
fitting [Nigmatullin et al., 2006].
The best criterion for melon quality is the
content of soluble solids [Dull et al., 1990],
which are mostly sugars (i.e., 80-85%),
and therefore a measure of sweetness. This
requires the extraction of tissue samples from
the melons and measurement of expressed juice
with a refractometer, which has been calibrated
to indicate percentage, by weight, of soluble
solids. In the studies on honeydew melons,
although correlations with soluble solids content
(SSC) were not judged sufficiently high, a high
correlation was found for a complex-planeplot of the dielectric constant and loss factor,
each divided by SSC. In subsequent research,
honeydew melons were grown and harvested
with a range of maturities, and dielectric
properties were measured for both the surface
and internal tissues along with SSC. Results of
the new studies are reported in this paper.
MATERIALS AND METHODS
Honeydew Melons
Three cultivars of honeydew melons (Cucumis
melo L. Inodorus group, ‘Early Dew’,
‘Honey Brew’, and ‘Rocio’) were grown
at the University of Georgia Horticulture
farm, Watkinsville, Georgia, USA, in a Cecil
clay loam soil using standard melon culture
methods [Boyham et al., 2006]. Plantings
were made on June 8, 13, and 19, 2006, with
seeds spaced 1 × 3 m. Fungicide applications
were alternated between Bravo and Pristine at
two-week intervals. Fertilization was based
upon soil analysis and culture recommendations
[Boyham et al., 2006]. Fruit of varying stages
of maturity was harvested over a 5-week period
for analysis. Honeydew melons not measured
on the harvest day were stored at 4°C from time
of harvest until time for measurements, which
was not more than a few days. Measurements
were taken on fourteen, twenty, and ten melons
of the ‘Early Dew’, ‘Honey Brew’, and ‘Rocio‘
cultivars, respectively.
Dielectric Properties Measurements
The electrical measurements necessary for
dielectric properties determination were
obtained with a Hewlett-Packard1 85070B openended coaxial-line probe, and a Hewlett-Packard
4291A Impedance/Material Analyzer [Nelson et
Mention of company or trade names is for purpose of description only and does not imply endorsement by the U. S.
Department of Agriculture.
1
International Microwave Power Institute
41-2-45
Figure 1. Surface measurement of honeydew
melon dielectric properties with open-ended
coaxial-line probe and impedance analyzer.
al., 2006]. Permittivities (dielectric constants
and loss factors) were calculated with Agilent
Technologies 85070D Dielectric Probe Kit
Software, modified for use with the HP 4291A
Analyzer by Innovative Measurement Solutions,
which provided permittivity values from the
reflection coefficient of the material in contact
with the active tip of the probe [Blackham and
Pollard, 1997]. Settings were made to provide
measurements at 51 frequencies on a logarithmic
scale from 10 MHz to 1.8 GHz. The 4291A
Analyzer was calibrated with an open, short, and
matched load prior to the calibration of the openended coaxial-line probe with measurements
on air, a short-circuit block, and glass-distilled
water at 25 °C. A personal computer was used to
control the system and record resulting data.
measurements were made with the probe in firm
contact with the surface of the melons in the
equatorial region at four points about 90° apart
around the perimeter of the melon. Melons were
supported and maintained in firm contact with the
probe with a laboratory jack as shown in Figure
1. Samples for measurements on the internal
melon tissue were all taken from an equatorial
slice, about 4 cm thick, cut with a sharp knife
from the center of the melon perpendicular to the
proximal-to-distal (stem to blossom) axis. From
this slice, three cylindrical core samples were
cut for the dielectric properties measurements as
shown in Figure 2. Three more samples adjacent
to the first samples were taken for moisture
content determination. Moisture contents were
determined by drying the triplicate samples in
disposable 57-mm aluminum weighing dishes
that were placed in a forced-air drying oven
for 24 hours at 70 °C. Upon removal from the
oven, weighing dishes with samples were cooled
in a desiccator over anhydrous CaSO4 prior to
reweighing for determination of moisture loss.
Samples were cut from the equatorial melon
slice with a cylindrical cutter [Nelson, 2003]
that provided right circular cylindrical samples
of 18.6-mm diameter and 3-to 4-cm length. All
permittivity measurements were taken at 24 °C.
Measurement procedures
Melons were removed from 4-°C storage the
night before measurements were scheduled and
washed with tap water to remove any dust or
soil material and permitted to equilibrate to 24
°C for the measurements. Initial permittivity
41-2-46
Figure 2. Equatorial slice of honeydew melon
and location from which cylindrical tissue
samples were cut.
Journal of Microwave Power & Electromagnetic Energy
Vol. 41, No. 2, 2007
of the sample. Thus, six series of permittivity
measurements were made for each tissue sample
at 51 frequencies from 10 MHz to 1.8 GHz to
be averaged for each melon. The four external
surface measurements were also averaged in a
similar way.
Upon completion of the internal tissue
permittivity measurements, each sample was
placed in a 1-oz. (30-ml) glass jar with screwon cap and held from a few minutes to 3 hours
for soluble solids content determination. The
measurements for soluble solids were determined
with an Atago Pallete Series Model PR101α
digital refractometer. Melon tissue samples were
7 placed in a garlic press with cheesecloth patches
Figure 3. Internal tissue sample measurement
of several layers to strain the juice expelled for
with open-ended coaxial-line probe.
the refractometer measurements. Five readings
were taken for each sample, and the composite
Samples
wereofsupported
as shown
in Figure
3 on than those of the surface measurements at the higher
dielectric
constants
the internal
tissues
are greater
soluble solids content determination for each
an aluminum platform that was raised to bring
melon was a mean of 15 readings.
the sample
into fiofrmthecontact
theare
openfrequencies.
Loss factors
internalwith
tissues
much higher than those of the surface measurements at the lower
ended coaxial-line probe for the permittivity
frequencies,
but they are[Nelson,
quite similar
at the
higher
measurements
2003].
When
thatend of the frequency range, thus illustrating the dominant
measurement was completed, the sample was
influence
of ionic conduction
theend,
lowerand
frequencies
and the dipolar losses at the higher frequencies.
removed,
turned endatfor
reinserted
for permittivity measurements on the other end
160
140
Permittivity
120
Dielectric constant
Loss factor
100
80
60
40
20
0
107
108
109
Frequency, Hz
Figure
44 melons.
melons.
Figure4.4.Mean
Meanpermittivity
permittivityvalues
valuesfor
forexternal
externalsurface
surface measurements
measurements on
on 44
International Microwave Power Institute
600
41-2-47
10
10
10
Frequency, Hz
Figure 4. Mean permittivity values for external surface measurements on 44 melons.
600
500
Permittivity
400
Dielectric constant
Loss factor
300
200
100
0
107
108
109
Frequency, Hz
Figure 5. Figure
Mean permittivity
values for values
internalfor
tissue
measurements
on 44 melonsonwith
mean moisture
content of
5. Mean permittivity
internal
tissue measurements
44 melons
with mean
91.3%, wet basis,moisture
and meancontent
solubleofsolids
content
of
8.4%
91.3%, wet basis, and mean soluble solids content of 8.4%.
RESULTS AND DISCUSSION
Frequency Dependence of Permittivities of
Honeydew Melons
Mean values for all the permittivity
measurements on the 44 honeydew melons are
shown in Figures 4 and 5. Permittivity values and
their frequency dependence are quite different
for the external surface and internal tissue
measurements. Although the dielectric constants
behave similarly with frequency, those for the
surface measurements are somewhat greater than
the internal tissue measurements at the lower
frequencies, and the dielectric constants of the
internal tissues are greater than those of the
surface measurements at the higher frequencies.
Loss factors of the internal tissues are much
higher than those of the surface measurements
at the lower frequencies, but they are quite
similar at the higher end of the frequency range,
thus illustrating the dominant influence of ionic
conduction at the lower frequencies and the
41-2-48
dipolar losses at the higher frequencies. When
the data from Figure 5 are observed in a log-log
plot, as shown in Figure 6, the linear relationship
between the log of the dielectric loss factor and
the log of frequency is evident at frequencies
below 600 MHz, which is indicative of the
dominance of ionic conductivity at the lower
frequencies. Similar results were noted for the
dielectric loss factors from tissue measurements
on other fruits and some vegetables [Nelson,
2005]. However, data for the external surface
permittivity measurements (Figure 4), reveal
an overriding dielectric relaxation behavior.
Judging from the frequency range for this
relaxation, the behavior is likely explained by
bound water and Maxwell-Wagner relaxations
[Hasted, 1973].
Correlation Between Permittivity and Soluble
Solids Content
To study the correlations of the dielectric
properties with the chosen quality factor, soluble
Journal of Microwave Power & Electromagnetic Energy
Vol. 41, No. 2, 2007
[Hasted, 1973].
.
1000
Permittivity
Dielectric constant
Loss factor
100
10
107
108
109
Frequency, Hz
Figure 6. Log-log plot of the permittivity of internal tissues of honeydew melons.
Figure 6. Log-log plot of the permittivity of internal tissues of honeydew melons.
solids content (SSC), linear regressions of the were obtained at 1.8 GHz.
dielectric constant ε ′ , loss factor ε ′′ and loss
The high correlation shown in Figures 8
Correlation Between Permittivity and Soluble Solids Content
tangent ( tan δ = ε ′′ / ε ′ ) on SSC were calculated and 9 leads one to expect that the dielectric
To study
correlations
of the
dielectric
with the
chosen quality
soluble
solidsthe
content
(SSC),
for the
the 51
frequencies
from
10 MHzproperties
to 1.8 GHz.
properties
might factor,
be used
to predict
soluble
Coefficients of determination, r2, for the dielectric
solids content [Nelson et al., 2006]. Referring to
linear constant
regressions
thefactor
dielectric
constant
� � surface
, loss factorFigure
� �� and
tangent
( tan
� � � �� /of
� the
) onstraight
SSC were
andof
loss
from the
external
8 orloss
Figure
9, the
equation
measurements were very low, less than 0.25 at line can be written as
calculated for the 51 frequencies from 10 MHz to 1.8 GHz. Coefficients of determination, r2, for the dielectric
all frequencies. Better correlations were found
 ε′

forand
the loss
dielectric
of the internal
ε ′′ low, less
constant
factor properties
from the external
surface tissue
measurements were very
(1)
= a f  −than
k  0.25 at all frequencies.
measurements, with highest r2 values at 1.8 GHz.
s
s


ε ′ and 0.752
r2 values
fordielectric
Better Those
correlations
were were
found0.675
for the
properties of the internal tissue measurements, with highest r2
for tan δ . The regression of tan δ on SSC is
where s represents soluble solids content, a f
illustrated in Figure 7.
In previous studies on honeydew melons is the slope of the line, and k is the ε ′ / s -axis
intercept. Values for the regression constants a f
[Nelson et al., 2006] a high correlation was
found between the permittivity and SSC in a and k are provided by the regression calculation.
complex-plane plot of ε ′′ / SSC and ε ′ / SSC at Therefore, solving (1) for s provides SSC in
terms of the dielectric properties as
1.8 GHz such as those shown in Figures 8 and 9.
Figure 8 shows results for the external surface
a f ε ′ − ε ′′
measurements at 1.8 GHz on mean values for the
s
=
(2)
44 honeydew melons in the work reported here,
af k
and Figure 9 shows results for the internal tissue
measurements. Linear regressions of ε ′′ / SSC
on ε ′ / SSC were calculated at all 51 frequencies For the melon data of Figure 8, a f = 0.2104 and
k = 0.2239. For that of Figure 9, a f = 0.1198
for both the external surface and internal tissue
and k = −3.4891.
measurements. Results are shown in Figure 10
where the best coefficients of determination
International Microwave Power Institute
41-2-49
0.21
0.20
Loss tangent
0.19
0.18
0.17
0.16
0.15
0.14
0.13
4
6
8
10
12
14
Soluble solids content, %
Figure 7. Linear regression of loss tangent on soluble solids content for internal tissues of
honeydew melons at 1.8 GHz, r2 = 0.752.
10
2.4
2.2
2.0
1.8
�''/SSC
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
2
4
6
8
10
12
�'/SSC
Figure 8. Complex-plane plot for 1.8-GHz permittivity from external surface measurements
Figure 8. Complex-plane
plot for
permittivity
from
surface
measurements
divided by soluble
divided
by 1.8-GHz
soluble solids
content
forexternal
honeydew
melons,
r2 = 0.954.
solids content for honeydew melons, r2 = 0.954.
2.4
41-2-50
2.2
2.0
Journal of Microwave Power & Electromagnetic Energy
Vol. 41, No. 2, 2007
�'/SSC
Figure 8. Complex-plane plot for 1.8-GHz permittivity from external surface measurements divided by soluble
solids content for honeydew melons, r2 = 0.954.
2.4
2.2
2.0
�''/SSC
1.8
1.6
1.4
1.2
1.0
0.8
4
6
8
10
12
14
16
�'/SSC
Figure 9. Complex-plane plot for 1.8-GHz permittivity from
11 internal tissue measurements divided by soluble solids
953. for 1.8-GHz permittivity from internal tissue measurements
content for
honeydew
melons, r2 = 0. plot
Figure
9. Complex-plane
divided by soluble solids content for honeydew melons, r2 = 0. 953.
1.0
Coefficient of determination
0.9
0.8
0.7
0.6
External surface measurements
Internal tissue measurements
0.5
0.4
0.3
0.2
107
108
109
Frequency, Hz
Fig.10. Coefficients of determination for linear regressions of � ' ' / SSC on � ' / SSC for 44 honeydew melons
Figure 10. Coefficients of determination for linear regressions of ε ′′ / SSC on ε ′ / SSC for 44
honeydew melons.
The high correlation shown in Figures 8 and 9 leads one to expect that the dielectric properties might be used to
Microwave Power Institute
41-2-51
predict theInternational
soluble solids
content [Nelson et al., 2006]. Referring to Figure 8 or Figure 9, the equation
of the straight
line can be written as
use the relationship illustrated in Figures 8 for more reliable prediction of SSC lies in the fact that neither
� � nor � ��
were individually well correlated with SSC.
14
Predicted SSC, %
12
10
8
6
4
2
2
4
6
8
10
12
14
Measured SSC,%
Fig. 11 Predicted SSC values from permittivity measurements on internal tissues of 44 honeydew melons, r2 = 0. 748.
Figure 11. Predicted SSC values from permittivity measurements on internal tissues of 44 honeydew melons, r2 = 0. 748.
PenetrationDepth
Soluble
content,
s, was calculated
Since thesolids
attenuation
of electromagnetic
waves in the melon tissue is important in attempting
 quality, the
2 to sense
 ε ′′ 
from Equation (2) for the 44 melons used in
2π ε ′ 
= attenuation
1is +described
1 the
  −by
nepers/m
penetration depth at the different frequencies is of interest. αThe
attenuation (3)

this study,
both for the external surface measureλ0 2 

 ε′ 


mentsconstant,
and the� internal
tissue
measurements,
, in terms of
the dielectric
properties of for
the material as follows:
comparison with measured SSC. Predicted SSC
where λ0 is the free-space wavelength. For
did not compare very well for the exterior surattenuation in decibels, dB/cm = 0.08686
face measurements, r2 = 0.015. For the internal
2
�
�
�
2� �
� ��
Therefore, the attenuation
to
� 1 � �� × ��nepers/m.
(3)
� 1� nepers/m
� better,
tissue measurements, predicted SSC �was
� � expected
2
�
�
�
�
�
0
be
in
honeydew
melon
tissue
was
�
�
r2 = 0.748 as illustrated in Figure 11. Thus the
calculated
with the measured dielectric constant
opportunity for a nondestructive determination
dB/cm
= 0.08686
� nepers/m.from
Therefore,
the
where �0 is the free-space wavelength. For attenuation in decibels,
and loss factor
data obtained
the external
from external surface measurements does not apsurface and internal tissue measurements
pear promising.
explanation
for the
inability
attenuation to The
be expected
in honeydew
melon
tissue was calculated with the measured dielectric constant and loss
to use the relationship illustrated in Figures 8 for all frequencies in the range used in these
factor data obtained from the external surface and internal tissue measurements for all frequencies in the range used
for more reliable prediction of SSC lies in the measurements. The depth of penetration,
d p=1/(2α), based on the average measured
fact that neither ε ′ nor ε ′′ were individually well
permittivities is shown in Figure 12. At the lower
correlated with SSC.
frequencies, the depth of penetration is much
greater, based on the permittivity values obtained
Penetration Depth
by the external surface probe measurements, than
it is in the internal melon tissue. At microwave
Since the attenuation of electromagnetic waves
in the melon tissue is important in attempting to frequencies, the indicated depth of penetration
is nearly the same for internal tissue and that near
sense quality, the penetration depth at the different frequencies is of interest. The attenuation the surface of the melons.
is described by the attenuation constant, α, in
terms of the dielectric properties of the material
as follows:
Correlation Between Moisture Content and
SSC
Correlations of dielectric constant and loss
factor with moisture content (wet basis) were
41-2-52
Journal of Microwave Power & Electromagnetic Energy
Vol. 41, No. 2, 2007
melons.
Penetration depth, cm
100
External surface measurements
Internal tissue measurements
10
1
107
108
109
Frequency, Hz
14 based on permittivities calculated from external surface
Figure 12. Dependence of penetration depth on frequency,
Figure 12. Dependence of penetration depth on frequency, based on permittivities calculated
probe measurements and probe measurements on internal tissues.
from external surface probe measurements and probe measurements on internal tissues.
96
Moisture content, %
94
Correlation Between Moisture Content
and SSC
92
Correlations of dielectric constant and loss factor with moisture content (wet basis) were low, r2 less than 0.2.
90
However, SSC was highly correlated
with tissue moisture content in honeydew melons [Nelson et al., 2006]. For
the 44 honeydew melons in this study, SSC was highly correlated with moisture content as shown in Figure 13.
88
86
4
6
8
10
12
14
SSC, %
Figure
13. Linear
Linearregression
regression
of moisture
content,
on soluble
solidss,content,
s, formelons,
honeydew
Fig. 13
of moisture
content,
M, on M,
soluble
solids content,
for honeydew
M � 99.5 � 0.985s , r2=0.972 melons, M=99.5-0.985s, r2=0.972.
low,CONCLUSIONS
r2 less than 0.2. However, SSC was highly
CONCLUSIONS
correlated
with
tissue
moisture
content
in
honMeasurements of dielectric constant and loss factor with an open-ended coaxial-line probe and impedance analyzer
eydew melons [Nelson et al., 2006]. For the 44
Measurements of dielectric constant and loss
on external surfaces and internal tissue of three cultivars of honeydew melons provided new permittivity data over a
honeydew melons in this study, SSC was highly
factor with an open-ended coaxial-line probe
correlated
with
moisture
content
as
shown
in
impedance
analyzer
on external
surfaces
range of maturities at frequencies from 10 MHz to 1.8 GHz atand
24 �C.
Both the dielectric
constant
and loss factor
of and
Figure 13.
internal tissues of three cultivars of honeydew
internal tissues decreased regularly with increasing frequency showing the dominance of ionic conduction at lower
melons provided new permittivity data over a
range
of maturities
from 10 MHz
frequencies and dipolar losses at the higher frequencies. Surface
measurements
revealedata frequencies
prominent and overriding
dielectric relaxation at lower frequencies and the characteristic decline in dielectric constant value with increasing
frequency. At all frequencies in the range studied, individual correlations of dielectric constant and loss factor with
International Microwave Power Institute
41-2-53
soluble solids content (SSC) (sweetness) were low, but coefficients of determination of about 0.7 were observed for
the loss factor and loss tangent of internal tissues at 1.8 GHz. However, the complex-plane plot of dielectric
to 1.8 GHz at 24 °C. Both the dielectric constant
and loss factor of internal tissues decreased
regularly with increasing frequency showing
the dominance of ionic conduction at lower
frequencies and dipolar losses at the higher
frequencies. Surface measurements revealed a
prominent and overriding dielectric relaxation
at lower frequencies and the characteristic
decline in dielectric constant value with
increasing frequency. At all frequencies in
the range studied, individual correlations of
dielectric constant and loss factor with soluble
solids content (SSC) (sweetness) were low, but
coefficients of determination of about 0.7 were
observed for the loss factor and loss tangent
of internal tissues at 1.8 GHz. However, the
complex-plane plot of dielectric constant and
loss factor, each divided by SSC, provided a high
correlation, r2 = 0.95 for both the external surface
and internal tissue measurements. Prediction of
SSC values from these relationships between
SSC and the dielectric properties, however, was
low with a coefficient of determination, r2, of
0.015 for surface measurements and 0.75 for
internal tissue measurements. Reliability of SSC
predictions from the dielectric properties was
low because of the poor individual correlations
of dielectric properties with SSC. Penetration
depths for electromagnetic fields in these
frequencies ranged from about 2 cm at the highest
frequencies to about 100 cm and 15 cm for the
surface and internal tissues, respectively, at 10
MHz. Further studies are needed to determine
whether practically useful techniques can be
developed for reliable prediction of quality from
dielectric properties of the melons.
REFERENCES
Blackham, D.V. and Pollard, R.D. 1997. An improved
technique for permittivity measurements using a coaxial probe. IEEE Transactions on Instrumentation
and Measurement 46(5): 1093-1099.
Boyham, G.E., Granberry, D.M. and Kelly, W.T. 2006.
Commercial Watermelon Production, University of
41-2-54
Georgia, College of Agricultural and Environmental
Sciences, http://pubs.caes.uga.edu/caespubs/pubed/
B996-w.htm.
Dull, G.G., Ressureccion, A.V.A. and Smittle, D.A. 1990.
Consumer panel acceptance of honeydew melons of
varing stages of maturity, 87th Annual Meeting,
Southern Association of Agricultural Scientists,
Food Sciences and Human Nutrition Section, Little
Rock, Arkansas, pp. 10.
Guo, W., Zhu, X. and Guo, K. 2001. Research Development of Nondestructive Determination Technologies
of Internal Qualities in Fruits. Transactions of the
CSAE 17(5): 1-5 (In Chinese).
Hasted, J.B. 1973. Aqueous Dielectrics. Chapman and
Hall, London.
Kays, S.J. and Paull, R.E. 2004. Postharvest Biology. Exon
Press, Athens, Georgia.
Nelson, S.O. 1980. Microwave dielectric properties of
fresh fruits and vegetables. Transactions of the ASAE
23(5): 1314-1317.
Nelson, S.O. 2003. Frequency- and temperature-dependent permittivities of fresh fruits and vegetables from
0.0l to 1.8 GHz. Transactions of the ASAE 46(2):
567-574.
Nelson, S.O. 2005. Dielectric spectroscopy of fresh fruit
and vegetable tissues from 10 to 1800 MHz. Journal
of Microwave Power and Electromagnetic Energy
40(1): 31-47.
Nelson, S.O., Forbus, W.R., Jr. and Lawrence, K.C. 1994.
Permittivities of fresh fruits and vegetables at 0.2 to
20 GHz. Journal of Microwave Power and Electromagnetic Energy 29(2): 81-93.
Nelson, S.O., Forbus, W.R., Jr. and Lawrence, K.C. 1995.
Assessment of microwave permittivity for sensing
peach maturity. Transactions of the ASAE 38(2):
579-585.
Nelson, S.O., Trabelsi, S. and Kays, S.J. 2006. Dielectric
spectroscopy of honeydew melons from 10 MHz
to 1.8 GHz for quality sensing. Transactions of the
ASABE 49(6): 1977-1981.
Nigmatullin, R.R., Arbuzov, A.A., Nelson, S.O. and
Trabelsi, S. 2006. Dielectric relaxation in complex
systems: quality sensing and dielectric properties of
honeydew melons from 10 MHz to 1.8 GHz. Journal
of Instrumentation 1(P10002): 1-19.
Seaman, R. and Seals, J. 1991. Fruit pulp and skin dielectric properties. Journal of Microwave Power and
Electromagnetic Energy 26(2): 72-81.
Journal of Microwave Power & Electromagnetic Energy
Vol. 41, No. 2, 2007