COLOR CHANGE ANALYSIS OF DRIED ORANGE SLICES DURING HOT AIR DRYING

Abstract

In this work, the hot air drying of orange slices was investigated in details, such as drying time, moisture ratio, dried products' colors and some macro-and micro-elements. The study aimed to apply a computer vision system to study the color changes due to drying. The orange slices were dried at hot air drying at temperatures of 70 °C and 80 °C and explored in detail such as drying time and mass degradations. Dried products of orange slices were subjected to color analysis by image analysis system in RGB color model. Three hundred color photographs of orange fruit slices taken before and after drying were analyzed.

Full text

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6064
COLOR CHANGE ANALYSIS OF DRIED ORANGE SLICES
DURING HOT AIR DRYING
Tunahan Erdem1,*, Omer Baris Ozluoymak1, Nacide Kizildag2
1Cukurova University, Department of Agricultural Machinery and Technologies Engineering, 01330, Adana, Turkey
2Cukurova University, Central Research Laboratory, 01330, Adana
ABSTRACT
In this work, the hot air drying of orange slices
was investigated in details, such as drying time,
moisture ratio, dried products’ colors and some
macro- and micro-elements. The study aimed to ap-
ply a computer vision system to study the color
changes due to drying. The orange slices were dried
at hot air drying at temperatures of 70 °C and 80 °C
and explored in detail such as drying time and mass
degradations. Dried products of orange slices were
subjected to color analysis by image analysis system
in RGB color model. Three hundred color photo-
graphs of orange fruit slices taken before and after
drying were analyzed.
KEYWORDS:
Drying, Image Analysis, Color Measurement, Orange.
INTRODUCTION
Drying is a widely used method to store the
fruit and vegetables for prolonged shelf life. Hot air
drying has been the utmost common industrial appli-
cation since the other technologies have higher in-
vestment fees. Therefore, the hot air drying was par-
ticularly preferred for this study and its effect on
color changes was identified by image analysis. To
preserve a fruit by drying causes some changes in
dried products. The term ‘product quality’ includes
three principal areas: nutritional value, acceptability,
and safety [1]. The effects of drying applications are
generally seen on the color of dried products. The
selection of high-quality products needs automation
to lower the labor and manufacturing costs. Nowa-
days, computer imaging is rapidly being developed
and image processing methods play an important
role in the development of food quality assessment
[2]. Usage of digital cameras and image processing
software have proven to be an ideal combination for
quick, cheap and accurate color measurement of var-
ious food products when compared to earlier tradi-
tional methods [3]. Computer imaging and image
processing could be used to classify the dried prod-
ucts based on color changes.
Earlier researches on color change have been
determined by either colorimeter or specific data ac-
quisition generally in L*a*b* units. Latterly, image
processing systems which provide some obvious ad-
vantages over a conventional colorimeter have been
developed to objectively measure the color of differ-
ent products [4]. Color space is defined by the CIE,
based on one channel for Luminance (lightness) (L)
and two channels for color (a and b). In this model,
the perceived color differences correspond to dis-
tances when measured calorimetrically. The axis ex-
tends from green (-a) to red (+a) and the axis b from
blue (-b) to yellow (+b). The brightness (L) increases
from the bottom to the top of the three-dimensional
model. On the other hand, RGB is the color model
utilized largely in display technologies that use light.
In this model, the colors red (R), green (G) and blue
(B) are added together at different intensities to pro-
duce millions of different colors on modern video
display screens.
Furthermore, some researcher have analyzed
color changes of certain food during drying, such as
kiwi [5], apple, banana, carrot and potato [6], tomato
cultivars [7], spinach [8], loquat fruit [9]. Moreover,
drying of orange slices has been studied by
Karaaslan and Erdem [10] as well as by Khafajeh et
al. [11]. In this study image of dried products and
fresh fruit were captured by digital camera and im-
age analysis were performed to determine the color
changes. Color deterioration of different product by
image analysis were also investigated by some re-
searcher such as; Yam et. al. [3], for pizza, Ze-
noozian et al. [12], for pumpkin, Faroogh et al. [1],
for fig fruit and Tarlak et al. [13], for raw meat, sa-
lami, sudjuk (Turkish style fermented sausage),
crushed red pepper flakes, cheddar cheese, potato,
biscuit, butter, (i) green bean, broccoli, white cheese,
cauliflower, black sesame and coconut flakes.
The objective of study was to evaluate the color
changes of dried orange slices by image analysis
method and fill the lack of knowledge in the research
area on it. The image analysis system could be
adopted the all production area from farm to fork to
eliminate the possible product losses and it decreases
the environmental pollution.

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6065
MATERIALS AND METHODS
Material. Citrus Sinensis (L.) Osbeck used in
this study were purchased from a local market. In or-
der to slow down the respiration, physiological and
chemical changes, they were stored at 4±0.5ºC be-
fore conducting the experiments [15]. Prior to dry-
ing, orange fruits were taken out of storage and they
were cut to 5 mm thick slices with a knife.
Method. Drying Procedure. The hot air dry-
ing was performed at the laboratory scale program-
mable oven. For the mass determination, a digital
balance of 0.01 g accuracy (Sartorius GP3202, Ger-
many) was used. Depending on the drying condi-
tions, moisture loss was recorded at 1-hour intervals.
The initial moisture content of the orange samples
was determined as 76.65% (w.b.) using a standard
method by drying in the oven at 105 ºC for 24 h. This
drying procedure was repeated three times.
Macro and Micro Element Analysis. Macro
and Micro Element Analyses were performed using
ICP-MS (Agilent, 7500a). The sample preparation
process was started by weighing 0.5 gr (dry mass) of
fruit sample into the digestion vessels. Then, 10 ml
of concentrated HNO3 (Merck) was added and the
mixture was decomposed in a CEM Mars 5 model
microwave (CEM Corporation Matthews, NC,
USA). Fruit samples were digested according to an
optimized program (power in W per minute:
1200/28, ventilation 10 min). During the final step
and the ventilation, the internal temperature was lim-
ited to 180 °C. After cooling, the digestion was trans-
ferred to 50 ml plastic bottles and diluted to 50 g with
twice deionized water and centrifuged at 4000 rpm
for 30 minutes. Reagent blanks were prepared, simi-
lar to the samples. All sample solutions were clear
and diluted 10 times prior to analysis.
Image Capturing System. An image captur-
ing system was designed and set up in a matte black
chamber in order to eliminate any environmental
light effects and a uniform light field was formed
around the sample. The size of the chamber was 300
mm (width) x 300 mm (length) x 300 mm (height) as
shown in Figure 1. The chamber material was se-
lected as matte black color to provide uniform reflec-
tance illumination onto the orange samples in the de-
tection area. The intensity of illumination on the sur-
face was measured at 1440 lx.
Using a color digital camera (Olympus E620,
Japan) images were captured through a hole on the
top surface. The calibrated camera settings for this
experiment were as follows: ISO:800; shutter-
speed:1/60; aperture:5.6; resolution:4032 x 3024;
format: JPEG. The distance between the lens and the
orange slices was 25 cm. The lighting system, com-
prising DC 12 V 48 LED lamps (Cata, TL-4481) was
placed around the camera lens upside of the box as a
square. A laptop computer (Acer, Aspire, 4830TG),
with an Intel Core i5 CPU and 8 GB RAM, was used
for the system software. Image processing process
was performed by using LabVIEW Professional De-
velopment and Vision Assistant Systems (National
Instruments; Austin, TX, USA). Totally 300 orange
slice images were subjected to image analyses for
fresh, dried by hot-air-drying at 70 and 80 °C
FIGURE 1
Image capturing system
1. CCD Camera 2. Chamber 3. LED lamps 4.
Sample
Digital Image Processing Method. In order to
extract some useful information or get an enhanced
image, digital image processing is a useful method
to convert an image into digital form and perform
some operations on it. In this study; firstly, histo-
gram analysis of fresh and dried orange slice images
was conducted in the LabVIEW, which is an object-
oriented graphical environment. A histogram is a
graph showing the frequency of repetition of color
values in an image. The graphic gives information
about whether the image is bright or dark. The histo-
gram of a digital image with intensity levels in the
range [0, L-1] is a discrete function
!
"
#$
%
&'$
[1]
where rk is the kth intensity value and nk is the num-
ber of pixels in the image with intensity rk [16].
With the help of the software, red (R), green
(G) and blue (B) components of an RGB image were
converted into a grayscale image. The mean value of
the histogram of an orange slice image was calcu-
lated automatically according to the dark pixels by
selecting the region of interest (ROI). The LabVIEW
user interface for the quantitative histogram analysis
is shown in Figure 2. As seen in Figure 2, only the
flesh of the orange slice was chosen to calculate the
mean value of the histogram.

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6066
FIGURE 2
LabVIEW user interface for the quantitative histogram analysis
FIGURE 3
Gray color index
FIGURE 4
The color changes of dried orange slices
In the dark images, the histogram was accumu-
lated in the low gray level region of the gray color
index; while in the bright images, the histogram was
accumulated in the high gray level region of the gray
color index (Fig. 3). Average color intensities for
both fresh and dried orange slices at 70°C and 80°C
were calculated from intensity histograms.
Secondly, the effect of temperature on color
changes of dried orange slices compared to the fresh
ones was investigated for 70°C and 80°C, respec-
tively. The color changes of dried orange slices were
calculated in RGB color space. Color features of
RGB images were extracted as red (R), green (G)
and blue (B) components. As studied in histogram
analysis, only the flesh of the orange slice was cho-
sen again to identify R, G and B values (Figure 4).
Then, RGB color values were converted into the CIE
L*, a*, b* color space to be analyzed for color
changes. RGB values were first converted to the
XYZ color space and then XYZ values were con-
verted to the CIE L*, a*, b* color space.
L*, a*, and b* components were calculated as
follows [1]:
()&**+
,-
.
./
0
1
2*+
[2]
3)&455
6-
7
7/
02
-
.
./
0
8 [3]
9)&:55
6-
.
./
022
-
;
;/
0
8 [4]

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6067
where (X0, Y0, Z0) were X, Y, Z values for standard
white, respectively. The value of X, Y and Z was
computed using a linear transformation from RGB
coordinates as follows:
,
7
.
;
1
&
<
5=+5> 5=*>? 5=:55
5=:@@ 5=4A> 5=**?
5=555 5=5++ *=**+
B,
C
D
E
1 [5]
In this coordinate system, the L* value is the
measure of lightness, ranging from 0 (black) to +100
(white); the a* value ranging from -100 (greenness)
to +100 (redness); and the b* value ranging from -
100 (blueness) to +100 (yellowness) [16].
And lastly, after the color values were ex-
pressed as L*, a* and b* before and after drying, to-
tal color change (ΔE), Chroma, hue angle and
browning index (BI) were calculated from the values
of L*, a* and b* and used to describe the color
change during drying. The Chroma or saturation in-
dex indicates color saturation and is proportional to
its intensity. The hue angle is another parameter fre-
quently used to characterize color in food products.
An angle of 0 or 2π radian represents red hue, while
angles of π/2, π, and 3π/2 radian represent yellow,
green and blue hues respectively [17]. The BI repre-
sents the purity of brown color. The total color
change (ΔE), Chroma, hue angle and BI were calcu-
lated from the following equations [1].
FG&
H"
(I
)2()
%
JK
"
3I
)23)
%
JK
"
9I
)29)
%
J
[6]
L!#MN3&
H
"3)J K9)J%
[7]
OPQR'STQ&UVWXY
Z
[)
\)
] [8]
^_&
`
YII
"
aXI=bY
%c
I=Yd ef&
"
\)gY=dhi)
%
"
h=jkhi)g\)Xb=IYJ[)
%
ll
[9]
where L*, a*, and b* values correspond to color val-
ues of orange slices at the end of drying and the val-
ues of Lo*, ao*, and bo* refer to the fresh orange
fruits. All experiments were performed in five repli-
cates and then averaged [1].
RESULTS AND DISCUSSION
Histogram Analysis Results. Histogram anal-
ysis results of fresh and dried orange slices after the
hot air drying are given in Figure 5. The grayscale
image was analyzed to find the probability of bright-
ness of fresh and dried orange slices. As seen in Fig-
ures 5, histogram values of dried orange slices de-
creased dramatically. Long-term drying applications
generally harm effect on the dried fruit. As it is ex-
pected, drying process at 70°C was carried out
longer than drying process at 80°C. As a result, or-
ange slices dried at 80°C had more favorable colors
than the dried at 70°C ones. In addition to that; aver-
age histogram graphs for fresh and dried orange
slices are given in Figure 6. The figure shows that
the better product could be easily separate from the
others by using image analysis tool.
FIGURE 5
Histogram analysis results of orange slices
FIGURE 6
Average histograms of gray color intensities of fresh and dried orange slices
90
111
133
154
175
1
11
21
31
41
51
61
71
81
91
101
Gray-level values of pixels
Histogram Value (Fresh) Histogram Value (70°C) Histogram Value (80°C)

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6068
FIGURE 7
Changes in the color parameters of orange slices
The Results of Color Parameters. The results
of color parameters are presented in Figure 7 for L*,
a* and b* values, respectively. As it is seen from fig-
ure, Low temperature of air at 70 °C caused to
browning of samples while had smallest values of
L*. Same results were reported by Pilli et al. [18].
They concluded this behavior could be due to long
treatment times. In contrast to this phenomenon, 80
°C drying application had the higher values of L*
40,00
45,00
50,00
55,00
60,00
65,00
70,00
75,00
80,00
85,00
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 101 106
L* Value
Sample
L*a*b* Value (Fresh) L* L*a*b* Value (70°C) L* L*a*b* Value (80°C) L*
0,00
5,00
10,00
15,00
20,00
25,00
30,00
35,00
40,00
111 21 31 41 51 61 71 81 91 101
a* Value
Sample
L*a*b* Value (Fresh) a* L*a*b* Value (70°C) a* L*a*b* Value (80°C) a*
40,00
45,00
50,00
55,00
60,00
65,00
70,00
75,00
80,00
85,00
111 21 31 41 51 61 71 81 91 101
b* Value
Sample
L*a*b* Value (Fresh) b* L*a*b* Value (70°C) b* L*a*b* Value (80°C) b*

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6069
due to the shortened drying time with compare to 70
°C drying application.
The figure illustrate that redness (a) value of
dried samples were higher than fresh one due to yel-
lowness (b) values were lower. Increasing the drying
temperature resulted decreased a* values. These data
are in contradict with earlier researcher report [18,
19].
Kinetic Parameters for the Color Change.
Kinetic parameters of color changes of orange slices
were investigated for fresh and dried orange slices.
By using L*, a* and b* values, total color change
(ΔE), chroma, hue angle and browning index (BI)
values were calculated before and after drying. Total
color change (ΔE) and average ΔE values before and
after drying of orange slices are given in Figure 8.
The other parameters such as Chroma, hue angle (ra-
dian) and browning index (BI) of orange slices be-
fore and after drying can be observed in Figure 9,
Figure 10 and Figure 11, respectively.
The figures indicate same results converted
from L*, a*, b* values. The conclusion is BI values
were different for each sample. As it is seen from
Figure 13 the 70 °C hot air application was highest
browning values. However 80 °C hot air application
values were close to fresh one.
FIGURE 8
Color changes and average ΔE values before and after drying
FIGURE 9
Average Chroma values of orange slices before and after drying
FIGURE 10
Average of hue angle values (rad) of orange slices before and after drying
40,4807
31,4389
12,2941
0,00
12,50
25,00
37,50
50,00
Average ΔE Values
ΔE (Fresh-70°C)
ΔE (Fresh-80°C)
ΔE (70°C-80°C)
77,2685
61,5687 63,121
0,00
20,00
40,00
60,00
80,00
100,00
Average Chroma Values
Chroma (Fresh) Chroma (70°C) Chroma (80°C)
1,49
1,05 1,17
0,00
0,50
1,00
1,50
2,00
Hue Angle (Radian)
Hue Angle (Fresh) Hue Angle (70°C) Hue Angle (80°C)

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6070
FIGURE 11
Average BI values of orange slices before and after drying
FIGURE 12
MC of dried orange slices
FIGURE 13
MR & drying time of dried orange slices
Drying Parameters of Orange Slices. Orange
slices were subjected to the hot air drying at 70 °C
and 80 °C. Drying parameters of dried samples can
be explored in MR and MC illustrated in Fig 12 and
13. From the figure, it is seen that the higher temper-
ature reduces the drying time 30 min as it is ex-
pected.
The reduced drying time affected the color
changes positively. As it is seen from Fig. 11 the
browning index shows the differences between 80
°C drying with fresh and 70 °C with dried samples.
All data from color analysis indicated the same re-
sults as the browning index. The color analysis re-
sults are not the only parameters to select the best
product but it is also needed to check the dried prod-
ucts for other quality parameters such as macro- and
micro-elements. Table 1 indicates versus effect on
macro- and micro-elements of dried samples. It
shows that the lower drying temperature has given
the higher amount of macro- and micro-elements.
The products’ sensitivity to temperature is a major
challenge in selecting the best method for drying.
218,5913
236,2633
215,9016
202,50
210,00
217,50
225,00
232,50
240,00
Browning Index
BI (Fresh) BI ( 70°C) BI (80°C)
0,00
1,00
2,00
3,00
4,00
5,00
6,00
0
30
60
90
120
150
180
210
240
270
300
330
360
390
MC kgH2O/kg[DM]
Drying Time (min)
80 °C HA ORANGE
70 °C HA ORANGE
0,00
0,20
0,40
0,60
0,80
1,00
1,20
0100 200 300 400
MR
Drying Time (min)
80 °C HA ORANGE
70 °C HA ORANGE

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6071
TABLE 1
Macro- and micro-element analysis on sliced orange fruit
Drying Type
C (%)
N (%)
P (%)
K (ppm)
Ca (ppm)
Mg (ppm)
Zn (ppm)
70oC
50,2
1,32
0,19
148
21,5
6,12
0,44
80oC
47,6
0,8
0,21
132
16,8
4,35
0,31
Sig
,000
,000
,070
,000
,000
,001
,000
*The datum were subjected to t test for equality.
The macro- and micro-element analysis results
were subjected to the t-test for equality in SPSS pro-
gram. C, N, K, Ca, Mg, Zn values were found sig-
nificantly different (sig. 0.000<0,05). That means,
orange slices had shown sensitivity to higher dying
temperatures in most macro- and micro-elements
due to versus effect seen in color.
CONCLUSION
As a conclusion, using the image analysis on
dried products for the purpose of selection promises
beneficial opportunities for automation systems
which computer vision system (CVS) provides an al-
ternative to the manual inspection of biological prod-
ucts by integrating an image acquisition device and
computer image analysis technique.
REFERENCES
[1] Sharifian, F., Motlagh, A.M., Komarizade,
M.H. and Nikbakht A.M. (2013) Colour Change
Analysis of Fig Fruit during microwave Drying.
International Journal of Food Engineering. 9(1),
107–114.
[2] Jelinski, T., Dub, C.J., Sun, D.W., Fornal, J.
(2007) Inspection of the distribution and amount
of ingredients in pasteurized cheese by com-
puter vision. Journal of Food Engineering. 83,
3–9.
[3] Yam, K.L., Papadakis, S.E. (2004) A simple
digital imaging method for measuring and ana-
lyzing color of food surfaces. J. Food Eng. 61,
137– 142.
[4] Brosnan, T., Sun, D. (2004) Improving quality
inspection of food products by computer vision
– a review. J Food Eng. 61, 3–16.
[5] Maskan, M. (2001) Kinetics of colour change of
kiwifruits during hot air and microwave drying.
Journal of Food Engineering. 48(2001) 169-
175.
[6] Krokida, M.K., Tsami, E., Maroulis, Z.B.,
(1998) Kinetics on Color Changes during Dry-
ing of Some Fruits and Vegetables. Drying
Technology. 16 (3-5), 667-685.
[7] Ashebir, D., Jezik, K., Weingartemann, H.,
Gretzmacher, R. (2009) Change in color and
other fruit quality characteristics of tomato cul-
tivars after hot-air drying at low final-moisture
content. International Journal of Food Sciences
and Nutrition. 60(S7), 308-315.
[8] Karaaslan, S.N., Tunçer I.K. (2008) Develop-
ment of a drying model for combined micro-
wave-fan-assisted convection drying of spinach.
Biosyst Eng. 100, 44–52.
[9] Wang, J., Lu, Z., Chen X., Zhang, H. (2016)
Modeling of Color Changes of Loquat Fruit dur-
ing Drying. Food Science. 37(21), 104-108.
[10] Karaaslan, S.N., Erdem, T. (2014) Mathemati-
cal Modelling of Orange Slices during Micro-
wave, Convection, Combined Microwave and
Convection Drying. Turkish Journal of Agricul-
tural and Natural Sciences. 1(2), 143–149.
[11] Khafajeh, H., Banakar, A, Ghobadian, B.,
Motevali, A. (2013) Drying of Orange Slices in
CHP Dryer. Advances in Environmental Biol-
ogy. 7(9), ISSN 1995-0756, 2326-2331.
[12] Zenoozian, M.S., Feng, H., Razavi, S.M.A.,
Shahidi, F., Pourreza, H.R. (2008) Image Anal-
ysis and Dynamic Modeling of Thin-Layer Dry-
ing of Osmotically Dehydrated Pumpkin. Jour-
nal of Food Processing and Preservation.
32(2008), 88–102.
[13] Tarlak, F., Ozdemir, M., Melikoglu, M. (2016)
Computer vision system approach in colour
measurements of foods: Part II. Validation of
methodology with real foods. Food Sci. Tech-
nol, Campinas. 36(3), 499-504.
[14] Maskan, M. (2001) Drying, shrinkage and rehy-
dration characteristics of kiwifruits during hot
air and microwave drying. Journal of Food En-
gineering. 48, 177-182.
[15] Gonzalez, R.C., Woods, R.E., (2008) Digital
Image Processing. Third Edition. Pearson Inter-
national Edition, Prentice Hall.
[16] Martynenko, A., Chen Y. (2013) Computer Vi-
sion for Real-Time Measurements of Shrinkage
and Color Changes in Blueberry Convective
Drying. Drying Technology. 31, 1114–1123.

© by PSP Volume 27 – No. 9/2018 pages 6064-6072 Fresenius Environmental Bulletin
6072
[17] Papadakis, S.E., Abdul-Malek, S., Kamdem,
R.E., Yam, K.L. (2000) A versatile and inexpen-
sive technique for measuring color of foods.
Food Technol. 2000(5), 48–51.
[18] Pilli, T., De Lovino, R., Maenza, S., Derossi, A.,
Severini, C. (2008) Study on Operating Condi-
tions of Orange Drying Processing: Comparison
Between Conventional and Combined Treat-
ment. Journal of Food Processing and Preserva-
tion. 32(2008), 751–769.
[19] Krokida, M.K., Maroulis, Z.B., Sarvacos, G.
(2001) The effect of the method of drying on the
colour of dehydrated products. Int. J. Food Sci.
Technol. 36, 53–59.
Received: 05.02.2018
Accepted: 13.06.2018
CORRESPONDING AUTHOR
Tunahan Erdem
Cukurova University,
Department of Agricultural Machinery and
Technologies Engineering,
01330, Adana – Turkey
e-mail: terdem@cu.edu.tr