Computers and Electronics in Agriculture 140 (2017) 311–316

Contents lists available at ScienceDirect

Computers and Electronics in Agriculture

journal homepage: www.elsevier.com/locate/compag

Original papers

Laser-induced backscattering imaging for classification of seeded and

seedless watermelons
Maimunah Mohd Ali a, Norhashila Hashim a,⇑, Siti Khairunniza Bejo a, Rosnah Shamsudin b
a
Department of Biological and Agricultural Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
b
Department of Food and Process Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: This paper evaluates the feasibility of laser-induced backscattering imaging for the classification of
Received 27 October 2016 seeded and seedless watermelons during storage. Backscattering images were obtained from seeded
Received in revised form 20 April 2017 and seedless watermelon samples through a laser diode emitting at 658 nm using a backscattering imag-
Accepted 10 June 2017
ing system developed for the purpose. The pre-processed datasets extracted from the backscattering
Available online 20 June 2017
images were analysed using principal component analysis (PCA). The datasets were separated into train-
ing (75%) and testing (25%) datasets as the inputs in the classification algorithms. Three multivariate pat-
Keywords:
tern recognition algorithms were used including linear discriminant analysis (LDA), quadratic
Backscattering imaging
Laser light
discriminant analysis (QDA), and k-nearest neighbour (kNN). The QDA-based algorithms obtained the
Watermelon highest overall average classification accuracies (100%) for both the seeded and seedless watermelons.
Pattern recognition algorithm The LDA and kNN-based algorithms also obtained quite high classification accuracies with all the accu-
Storage racies above 90%. The laser-induced backscattering imaging technique is potentially useful for classifica-
tion of seeded and seedless watermelons.
Ó 2017 Elsevier B.V. All rights reserved.

1. Introduction (Perkins-Veazie and Collins, 2004; Rimando and Perkins-Veazie,

Watermelon (Citrullus lanatus) is an edible and beneficial fruit
which is available all year round in Malaysia (Mohd Ali et al.,
2017). The fruit is the world’s second most famous fruit with
approximately 109.28 million metric tons global production in
2013 (FAO, 2015). Watermelon is well-known for its sweet and
juicy flesh. The fruit is denoted as a pepo, which is a special type
of berry that has a thick rind and deep flesh inside (Ajuru and
Okoli, 2013). The pepo is also developed from the inferior fruit
which is distinctive for the watermelon crop. Normally, the colour
of the internal watermelon flesh of most commercial cultivars is
red or yellow, while the watermelon rind colour varies from light
to dark green with yellow spots, stripes or marbling. As an added
value, the entire fruit is edible and is frequently eaten raw or con-
sumed as a juice, stir-fried, in a salad, or pickled.
Apart from the above, watermelon differs in terms of shape and
size from circular to oblong. The fruit has a high water content and
is a good source of vitamins and minerals such as carbohydrates,
calcium, phosphorus, and ascorbic acid (Tlili et al., 2011). Water-
melon is also proven to be a source of lycopene and citrulline
⇑ Corresponding author.
E-mail address: norhashila@upm.edu.my (N. Hashim).
2005). Citrulline is a non-essential amino acid obtained from the
watermelon juice whereas lycopene is a type of carotenoid that
gives the distinct property of a red colour (Liu et al., 2010). In addi-
tion, watermelon is a non-climacteric fruit which does not undergo
a continuous process to ripen after being harvested (Wechter et al.,
2008). The fruit belongs to the family of Cucurbitaceae that consists
of numerous cultivars, including seeded, seedless and disorder-
resistant varieties (Perkins-Veazie et al., 2006).
Technically, since the seedless watermelon cannot be planted
through seed germination, the fruit has to be cultivated by using
different scientific methods. The seeded watermelon is normally
pollinated using pollen, whereas the seedless watermelon is pro-
duced from a hybrid watermelon which consists of three sets of
chromosomes in terms of a genetically-engineered procedure
(Griffiths et al., 1999). However, as farmers start to plant seedless
cultivars, roughly one-third of the crops are left to plant seeded
watermelons for sufficient pollination and yield production.
Watermelon cultivars also like dry conditions, especially with
exposure to sunlight and temperature during the growing period.
There are several criteria to be considered when choosing a
variety of watermelon. The main characteristics are colour and
visual inspection of the fruit (Zhang et al., 2014). The harvesting
process is much easier for the seedless cultivar due to the different
colour compared to the pollination of the seeded variety. Size is

http://dx.doi.org/10.1016/j.compag.2017.06.010
0168-1699/Ó 2017 Elsevier B.V. All rights reserved.
312 M. Mohd Ali et al. / Computers and Electronics in Agriculture 140 (2017) 311–316
also one of the criteria as the shape and size vary between both
varieties (Zheng et al., 2006). Conversely, seedless watermelon is
grown to have immature seed which is developed from the triploid
trait and comprises of numerous large seeds throughout the flesh
of the hybrid watermelon. For this reason, the seedless variety is
more valuable in commercial markets due to the preference and
convenience requirements of the customers.
Recently, laser-induced backscattering imaging has been
applied for non-destructively evaluating various agricultural pro-
duce in terms of food quality and control as an alternative to the
machine vision method. Previous works have discussed the feasi-
bility of laser-induced backscattering imaging in evaluating the
food quality of fruits and vegetables such as apples (Qing et al.,
2007; Romano et al., 2011), bell peppers (Romano et al., 2012),
tomatoes (Mollazade et al., 2013; Tu et al., 2000), bananas
(Adebayo et al., 2016b; Hashim et al., 2014, 2013, 2012), kiwifruits
(Baranyai and Zude, 2009), citrus (Lorente et al., 2015; Lorente
et al., 2013), and papayas (Udomkun et al., 2014). Romano et al.
(2011) applied laser light backscattering imaging emitted in the
visible wavelength at 635 nm to determine moisture content, sol-
uble solids content, and hardness of apples during drying.
Lorente et al. (2013) used a laser diode emitting at five different
wavelengths (532, 660, 785, 830, and 1060 nm) to detect decay
in citrus fruits in comparison with respective laser wavelengths.
In previous research, Hashim et al. (2013) investigated the effect
of water content, fruit pigment, and chilling injury in bananas with
respect to storage time and temperature using laser-induced
backscattering imaging. Dénes et al. (2013) reported the applica-
tion of a backscattering imaging system on the optical properties
of bananas based on the influence of different drying times, tem-
peratures, and pre-treatments emitting at seven wavelengths
(532, 635, 650, 780, 808, 850, and 1064 nm). Qing et al. (2008)
analysed backscattering images of apples to determine the firm-
ness and soluble solids content in different growing locations.
Basically, laser induced backscattering imaging uses the theory
of light propagation when the light is reflected and scattered back
towards the external surface of the sample upon entering the fruit
tissue. In this case, the backscattering photons obtained through
light propagation carry useful information associated with the
morphology properties of the flesh when interacting with the qual-
ity attributes of the fruit sample (Adebayo et al., 2016a; Mollazade
et al., 2012). Moreover, the methods developed for processing the
backscattering images are computationally inexpensive. The low
processing cost suggests that a laser-based system could possibly
be developed and applied, especially in real-time sorting/grading
machines. However, as far as is known, no study has reported a
technique able to classify seeded and seedless watermelons cou-
pled with classifier models using a laser-induced backscattering
imaging system. Thus, the ultimate aim of this study is to evaluate
the application of laser-induced backscattering imaging for the
classification of seeded and seedless watermelons during storage.
This study also investigates different pattern recognition algorithm
models for the classification of seeded and seedless watermelons.
2. Materials and methods
2.1. Sample preparation
Seeded watermelons from the Black Beauty variety and seedless
watermelons from the Red Seedless variety were procured from a
local farm in Manjung, Perak, a middle state in Malaysia. The
watermelons were harvested 65 days after planting and delivered
to a commercial retailer (Melon Master Sdn. Bhd.). The fruits were
transported immediately to the Biomaterials Processing Labora-
tory, Universiti Putra Malaysia and stored in cold storage (12 °C
and 80–90% RH). The experiments were carried out in two batches
for three weeks in order to identify the similarity and differentia-
tion throughout the storage period. For each cultivar, 70 Black
Beauty and Red Seedless watermelons were measured with a total
of 140 watermelons for each batch, respectively. Seven interval
storage days were used to analyse the physicochemical changes
of both varieties (day 0, day 4, day 8, day 12, day 15, day 18, and
day 21). Ten watermelons were randomly sampled from both the
seeded and seedless watermelons for each storage day under the
same storage condition. The same fruits were also used for both
backscattering acquisition and destructive measurement.
2.2. Imaging system
In this study, a laser-induced backscattering imaging system
which consists of a charge-coupled device (CCD) camera (QICAM
Colour Fast 1394, QImaging, Surrey, BC, Canada) with a zoom lens
(F5.6 and focal length of 18 mm), a laser diode emitting at 658 nm
with 30 mW maximal power, and a computer equipped with the
Image-Pro Insight 9 software (Media Cybernetics, Inc., USA) was
developed for acquiring the backscattering images. The CCD cam-
era captured backscattering images of 1392  1040 pixels with a
resolution of 0.073 mm/pixel and 12-bit gray colour depth. The
portion of backscattering light upon penetrating into the fruit tis-
sue was recorded by the CCD camera and stored on the computer.
A frame (930 mm  700 mm  700 mm) with a lightproof medium
i.e. black cloth was used to avoid interruption from ambient light.
A laser diode at 658 nm is used as the light source which can dis-
tribute light for a selected wavelength. Since the processing of the
whole laser-induced backscattering imaging system is quite fast,
the image acquisition was set at a short exposure time (10 ms).
The distance from the camera lens to the samples was set at
55 cm in a fixed position at the top of the fruit. The incident angle
of the light beam was placed at 22° with respect to the vertical
axis. The backscattering images were obtained by placing the sam-
ple manually on the sample holder facing the CCD camera.
Backscattering images obtained at 658 nm from seeded and seed-
less watermelon samples during storage are shown in Fig. 1. Over-
all, a total of six images were acquired for each of the 140 samples
of watermelon skin, thereby resulting in 840 backscattering
images.
2.3. Backscattering image analysis
The backscattering images were obtained through an image
segmentation process using Matlab software (Version R2013a,
The Mathworks Inc., Natick, MA, USA). The region of interest in
the image was selected by choosing the threshold value from a his-
togram profile based on the backscattering image (Fig. 2). From the
feature extraction of the backscattering image, the pixel value
(maximum intensity, minimum intensity, and mean intensity)
and shape (major axis length, minor axis length, and perimeter)
measurements were determined. The backscattering parameter
measurements were then exported to Unscrambler software (Ver-
sion 10.3, CAMO AS, Oslo, Norway) for further statistical analysis.
2.4. Multivariate pattern recognition algorithms
Linear discriminant analysis (LDA), quadratic discriminant anal-
ysis (QDA), and k-nearest neighbour (kNN) algorithms are different
multivariate pattern recognition algorithms that can be used to
classify the watermelon cultivars and storage days based on the
backscattering parameters. These algorithms have been reported
in the literature for different classification purposes, especially to
classify spectral data sets (Liaghat et al., 2014; Sankaran et al.,
2011). Before determining the overall classification accuracies of
 M. Mohd Ali et al. / Computers and Electronics in Agriculture 140 (2017) 311–316
Fig. 1. Backscattering images of the fruit samples using laser diode with wavelength of 658 nm for (a) seeded and (b) seedless watermelons.
Fig. 2. Backscattering image analysis: (a) raw backscattering image and (b) histogram profile of pixels of image.
the different algorithms, the ratio of training to testing dataset
(70:30) was calculated using the combined datasets. The PC scores
were randomised and divided into training datasets for developing
the classification model whereas the testing datasets were used to
test the model. In this study, LDA and QDA used two PC scores
obtained from the combined datasets. The kNN model consisted
of two classes which were derived from the training datasets and
predicted by determining the distance between the unknown sam-
ple and k nearest neighbours. All multivariate algorithms were
developed in the SAS software (Version 9.4, SAS Institute, Cary,
NC, USA). The average classification accuracies were calculated
by taking the average of five readings to randomise the data sets.
The overall and individual classes (seeded and seedless) classifica-
tion accuracies were obtained from a confusion matrix.
Discriminant functions were obtained from the input indepen-
dent variables via the training data sets, which were used to deter-
mine the classification based on the principal component (PC)
score values between the classes. In discriminant analysis, LDA is
a linear model whereas QDA is known as a quadratic model, which
generates the classification of the testing datasets. Apart from the
discriminant analysis algorithms, a kNN algorithm was also used to
classify between the classes from the training datasets. The two
classes of datasets (seeded and seedless) were predicted by calcu-
lating the distance between the k nearest neighbours and the
classes of samples. The optimum k-value (constant) in the algo-
rithm was chosen from the highest classification accuracies during
the kNN algorithm development. In this case, the kNN algorithm
313
along with k-value yielded the highest overall and individual class
for the average classification accuracies and was regarded as the
best model for the kNN algorithm.
3. Results and discussion
3.1. Potential of laser-induced backscattering imaging to classify
watermelons
Principal component analysis (PCA) was performed to further
minimise the backscattering parameters into uncorrelated vari-
ables, also known as PC. A clear discrimination between the seeded
and seedless watermelons by PC1 and PC2 based on the PCA score
plot was observed (Fig. 3). Two PCs including PC1 captured 61% of
the variance and PC2 captured 27% of the variance thus showing a
clear separation between the seeded and seedless watermelons.
Based on the results, laser-induced backscattering imaging was
found to be able to provide the necessary classification of seeded
and seedless watermelons. The PC score was randomised and plot-
ted in an effort to classify the classes of correlated variables of the
seeded and seedless watermelons. A preliminary set of analyses
was performed on the samples where 75% of the datasets were
used for training the classification algorithm, and 25% of the data-
sets were used as the testing dataset.
By observing the PCA score plot, it was possible to classify the
relationship between the backscattering parameters as well as
314 M. Mohd Ali et al. / Computers and Electronics in Agriculture 140 (2017) 311–316

and kNN-based algorithms due to the deterioration of the fruit

Fig. 3. PCA score plot of seeded and seedless watermelons based on the measure-
ment of backscattering parameters (j Seeded; d Seedless).
the seeded and seedless watermelons. Extra work needs to be
undertaken to maximise the PCA classification model by consider-
ing the scattering and absorption coefficients acquired from the
backscattering images. The backscattering parameters were fur-
ther validated using these coefficients to create the classification
accuracy algorithm. Hence, the laser-induced backscattering imag-
ing technique can be used to successfully classify the seeded and
seedless watermelons based on the backscattering parameter mea-
surement. In addition, the multivariate statistical analysis regard-
ing PCA demonstrated reliable information and relevant results
for the explanation of the backscattering parameter
measurements.
3.2. Classification using LDA, QDA, and kNN
The average classification accuracies of the seeded and seedless
watermelons at different storage days using multivariate pattern
recognition algorithms are presented in Table 1. Based on the
results, the LDA-based algorithms yielded a lower average classifi-
cation accuracy than QDA and kNN, especially for the seedless
watermelon at day 12 (40%) and day 21 (30%). In comparison with
the three multivariate pattern recognition algorithms, it was found
that the QDA-based algorithm yielded a perfect score of average
classification accuracies (100%) for all seven storage days of the
seeded and seedless watermelons.
The kNN-based algorithms also yielded quite high classification
accuracies with results all above 60%, except for day 12 (50%) and
day 15 (50%) for the classification of seedless watermelon. For the
classification of seedless watermelon, the average classification
accuracies decreased along with the storage day for both LDA
when ripening. Since the severity worsened, the classification of
both cultivars became difficult and decreased as the postharvest
storage increased. Thus, it could be signified that QDA-based algo-
rithms performed as the most reliable algorithm to classify
between seeded and seedless watermelons over different storage
days.
The classification accuracies of seeded and seedless watermel-
ons over different storage days based on the backscattering param-
eters (major axis length, minor axis length, mean intensity,
maximum intensity, minimum intensity, and perimeter) were per-
formed using the QDA-based algorithm in order to find the correct
classification and miscalculation error (Table 2). The effectiveness
of the backscattering parameters for classifying between the
seeded and seedless watermelons was evaluated based on the stor-
age day. The classification accuracies and misclassification errors
analysis was performed on seven storage days using the QDA algo-
rithm since it shows the best classifier, tested on the overall data-
set samples. The correct classification and miscalculation errors
were expressed as a percentage (%) value according to the storage
day. Based on the findings, the correct classification varied among
the backscattering parameters for both seeded and seedless water-
melons. No similar trend could be found for each of the backscat-
tering parameters with respect to storage day. As shown in
Table 2, the classification error decreased as the storage day
increased. The percentage decline of the miscalculation error might
be due to the fruit ripening since the shelf-life of watermelon nor-
mally lasts for 21 days upon harvest. Thus, the changes of physic-
ochemical properties can be identified and classified accordingly
when subjected to backscattering analysis and the parameters
determined.
Among all the backscattering parameters, the maximum inten-
sity achieved the highest correct classification (100%) for three
storage days; day 4, day 12, and day 18. The mean intensity param-
eter also attained the highest correct classification (100%) for day 4
and day 18 for both the seeded and seedless watermelons. On the
other hand, the minor axis length achieved the highest miscalcula-
tion error (60%) for day 4 and day 18 in order to classify seeded
watermelon, respectively. In this case, no similar trend was
observed for the classification of seeded and seedless watermelons
since there was variation within each backscattering day and dif-
ferent storage days. However, the backscattering parameters were
found to be reliable to classify the seeded and seedless watermel-
ons based on the storage day.
3.3. Comparison of classification analysis
The average overall, seeded and seedless classification accura-
cies of three multivariate pattern recognition algorithms (LDA,
QDA, and kNN) are summarised in Fig. 4. Among the three different
multivariate pattern recognition algorithms, the QDA-based algo-
rithms achieved perfect and highest scores overall, for both seeded
and seedless classification accuracies (100%) to classify between

Table 1
The average classification accuracies of seeded and seedless watermelons at different storage days using multivariate pattern recognition algorithms.

Storage day LDA QDA kNN

Seeded Seedless Seeded Seedless Seeded Seedless
Day 0 60 50 100 100 75 60
Day 4 70 100 100 100 80 100
Day 8 70 70 100 100 67 60
Day 12 60 40 100 100 75 50
Day 15 70 60 100 100 75 50
Day 18 100 70 100 100 100 80
Day 21 100 30 100 100 100 75
 M. Mohd Ali et al. / Computers and Electronics in Agriculture 140 (2017) 311–316 315

Table 2
The classification accuracies of seeded and seedless watermelons at different storage days based on the backscattering parameters.

Storage Cultivar Major axis length Minor axis length Mean intensity Maximum intensity Minimum intensity Perimeter
day
Correct Missed Correct Missed Correct Missed Correct Missed Correct Missed Correct Missed
(%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) (%)
Day 0 Seeded 70 30 50 50 50 50 80 20 50 50 50 50
Seedless 80 20 80 20 70 30 100 0 70 30 90 10
Day 4 Seeded 70 30 60 40 100 0 100 0 80 20 50 50
Seedless 100 0 40 60 100 0 100 0 80 20 80 20
Day 8 Seeded 50 50 50 50 100 0 90 10 60 40 50 50
Seedless 100 0 100 0 60 40 70 30 100 0 90 10
Day 12 Seeded 60 40 50 50 100 0 100 0 60 40 50 50
Seedless 90 10 80 20 90 10 100 0 80 20 100 0
Day 15 Seeded 80 20 80 20 90 10 80 20 70 30 70 30
Seedless 90 10 90 10 70 30 70 30 100 0 90 10
Day 18 Seeded 80 20 40 60 100 0 100 0 100 0 60 40
Seedless 80 20 70 30 100 0 100 0 50 50 70 30
Day 21 Seeded 100 0 100 0 80 20 40 60 100 0 100 0
Seedless 100 0 100 0 70 30 90 10 100 0 100 0

and infected oil palm leaves using pre-processed Fourier transform

infrared (FT-IR) spectral data. As reported by Lorente et al. (2015),
the findings indicated that the LDA classifier acts as a function for
each feature selection in the early detection of decay in citrus fruit
using the Gaussian–Lorentzian cross product (GL) model and Far-
rell’s model with the overall classification accuracies of 93.38%
and 93.39%, respectively.

4. Conclusions

The potential of a laser-induced backscattering imaging system

Fig. 4. Average classification accuracies using the multivariate classification model.
the two watermelon cultivars. Meanwhile, the LDA and kNN-based
algorithms yielded quite high classification accuracies with all the
accuracies above 90%. In comparison of the three classification
accuracies, the kNN-based algorithm also received a perfect score
of 100% for the seeded classification accuracy. Since it is quite chal-
lenging to visually classify between the seeded and seedless water-
melons during postharvest storage, it is advisable to classify the
fruit in terms of the type of cultivar for postharvest handling.
Therefore, the QDA-based algorithm could be defined as the best
classifier model for the classification of seeded and seedless water-
melons. Hashim et al. (2013) used QDA to identify chilling injury in
bananas based on the treatment temperature and visual assess-
ment. The results indicated that the lowest misclassification error
was obtained at ripening stage 5 of bananas with a percentage of
error values of 4.33% and 1.67% during and after storage at
660 nm using a laser light backscattering imaging system.
However, there was no significant difference between the LDA
and kNN-based algorithms in the overall and seedless classifica-
tion. The interaction effect of the significant factors obtained from
the three models proved that the QDA-based algorithm showed the
highest classification accuracies in seeded watermelons despite the
slight difference between the classifications of the overall as well
as seedless watermelons. Similarly, Sankaran et al. (2011) reported
that the QDA can discriminate between healthy and infected citrus
leaves with high average overall classification accuracies of greater
than 90%. In previous research by Liaghat et al. (2014), the results
showed that LDA-based models gave the highest average overall
classification accuracies of 92% to distinguish between healthy
for classification of seeded and seedless was evaluated. The find-
ings highlighted that the three pattern recognition algorithms,
namely LDA, QDA, and kNN could classify between seeded and
seedless watermelons with high average classification accuracies
of greater than 90%. Based on the backscattering parameter mea-
surements, the QDA-based algorithm achieved perfect overall clas-
sification accuracies of 100% for both seeded and seedless
watermelons. The feasibility of laser-induced backscattering imag-
ing offers a new optical approach for classification of watermelons,
especially during the process of fruit grading/sorting. It is notewor-
thy to mention that the classification of the seeded and seedless
watermelons depending on visual inspection without cutting the
fruit in the field was difficult, therefore, needing an automated sys-
tem to identify between the two cultivars. This implies the require-
ment for developing a computerised device as an important factor
of potential success in the further classification of the seeded and
seedless watermelons by means of an automated system.
Although this study evaluated the potential of a laser-induced
backscattering imaging system for watermelon classification, a
wide range of data is required to further validate the existing
results. With future work of extending this study to the different
maturity levels of watermelons, an on-line sorting/grading
machine system for the fruit is a realistic goal with the ongoing
improvement in hardware features in laser-induced backscattering
imaging systems. The implementation of the system is also possi-
ble to be automated in sorting/grading machines because this tech-
nique requires a fast processing time for the backscattering images.
With the advancing progress of imaging technologies, the laser-
induced backscattering imaging system is suitable for automated
sorting/grading machines which can go beyond limited human
inspection to assess huge amounts of fruits to be inspected auto-
matically. The expansion of rapid automated systems, along with
316 M. Mohd Ali et al. / Computers and Electronics in Agriculture 140 (2017) 311–316
the development of suitable imaging techniques offers new and
robust devices by means of an automatic inspection system.
Acknowledgments
The authors are thankful for the support received from the
Department of Biological and Agricultural Engineering, Faculty of
Engineering, Universiti Putra Malaysia for supplying the facilities
for this research under the Fundamental Research Grant Scheme
(FRGS) (Vot. No.: 5524604).
References
Adebayo, S.E., Hashim, N., Abdan, K., Hanafi, M., 2016a. Application and potential of
backscattering imaging techniques in agricultural and food processing – a
review. J. Food Eng. 169, 155–164.
Adebayo, S.E., Hashim, N., Abdan, K., Hanafi, M., Mollazade, K., 2016b. Prediction of
quality attributes and ripeness classification of bananas using optical
properties. Sci. Hortic. (Amsterdam) 212, 171–182.
Ajuru, M.G., Okoli, B.E., 2013. The morphological characterization of the melon
species in the family Cucurbitaceae Juss., and their utilization in Nigeria. Int. J.
Mod. Bot. 3, 15–19.
Baranyai, L., Zude, M., 2009. Analysis of laser light propagation in kiwifruit using
backscattering imaging and Monte Carlo simulation. Comput. Electron. Agric.
69, 33–39.
Dénes, D.L., Parrag, V., Felföldi, J., Baranyai, L., 2013. Influence of parameters of
drying on laser induced diffuse reflectance of banana discs. J. Food Phys. 26, 11–
16.
FAO, 2015. Global production of vegetables in 2013, by type (in million metric tons).
Retrieved from <http://www.statista.com/statistics/264065/global-production-
of-vegetables-by-type/> (accessed 24.04.16).
Griffiths, A.J., Gelbart, W.M., Miller, J.H., Lewontin, R.C., 1999. Changes in
Chromosome Number. W. H. Freeman, New York.
Hashim, N., Janius, R.B., Abdul, R., Osman, A., Shitan, M., Zude, M., 2014. Changes of
backscattering parameters during chilling injury in bananas. J. Eng. Sci. Technol.
9, 314–325.
Hashim, N., Janius, R.B., Baranyai, L., Rahman, R.A., Osman, A., Zude, M., 2012.
Kinetic model for colour changes in bananas during the appearance of chilling
injury symptoms. Food Bioprocess Technol. 5, 2952–2963.
Hashim, N., Pflanz, M., Regen, C., Janius, R.B., Abdul Rahman, R., Osman, A., Shitan,
M., Zude, M., 2013. An approach for monitoring the chilling injury appearance
in bananas by means of backscattering imaging. J. Food Eng. 116, 28–36.
Liaghat, S., Mansor, S., Ehsani, R., Shafri, H.Z.M., Meon, S., Sankaran, S., 2014. Mid-
infrared spectroscopy for early detection of basal stem rot disease in oil palm.
Comput. Electron. Agric. 101, 48–54.
Liu, W., Zhao, S., Cheng, Z., Wan, X., Yan, Z., King, S.R., 2010. Lycopene and citrulline
contents in watermelon (Citrullus lanatus) fruit with different ploidy and
changes during fruit development. Acta Hortic. 871, 543–550.
Lorente, D., Zude, M., Idler, C., Gómez-Sanchis, J., Blasco, J., 2015. Laser-light
backscattering imaging for early decay detection in citrus fruit using both a
statistical and a physical model. J. Food Eng. 154, 76–85.
Lorente, D., Zude, M., Regen, C., Palou, L., Gómez-Sanchis, J., Blasco, J., 2013. Early
decay detection in citrus fruit using laser-light backscattering imaging.
Postharvest Biol. Technol. 86, 424–430.
Mohd Ali, M., Hashim, N., Bejo, S.K., Shamsudin, R., 2017. Quality evaluation of
watermelon using laser-induced backscattering imaging during storage.
Postharvest Biol. Technol. 123, 51–59.
Mollazade, K., Omid, M., Akhlaghian Tab, F., Kalaj, Y.R., Mohtasebi, S.S., Zude, M.,
2013. Analysis of texture-based features for predicting mechanical properties of
horticultural products by laser light backscattering imaging. Comput. Electron.
Agric. 98, 34–45.
Mollazade, K., Omid, M., Tab, F.A., Mohtasebi, S.S., 2012. Principles and applications
of light backscattering imaging in quality evaluation of agro-food products: a
review. Food Bioprocess Technol. 5, 1465–1485.
Perkins-Veazie, P., Collins, J.K., 2004. Flesh quality and lycopene stability of fresh-
cut watermelon. Postharvest Biol. Technol. 31, 159–166.
Perkins-Veazie, P., Collins, J.K., Davis, A.R., Roberts, W., 2006. Carotenoid content of
50 watermelon cultivars. J. Agric. Food Chem. 54, 2593–2597.
Qing, Z., Ji, B., Zude, M., 2008. Non-destructive analyses of apple quality parameters
by means of laser-induced light backscattering imaging. Postharvest Biol.
Technol. 48, 215–222.
Qing, Z., Ji, B., Zude, M., 2007. Predicting soluble solid content and firmness in apple
fruit by means of laser light backscattering image analysis. J. Food Eng. 82, 58–
67.
Rimando, A.M., Perkins-Veazie, P.M., 2005. Determination of citrulline in
watermelon rind. J. Chromatogr. A 1078, 196–200.
Romano, G., Argyropoulos, D., Nagle, M., Khan, M.T., Müller, J., 2012. Combination of
digital images and laser light to predict moisture content and color of bell
pepper simultaneously during drying. J. Food Eng. 109, 438–448.
Romano, G., Nagle, M., Argyropoulos, D., Müller, J., 2011. Laser light backscattering
to monitor moisture content, soluble solid content and hardness of apple tissue
during drying. J. Food Eng. 104, 657–662.
Sankaran, S., Mishra, A., Maja, J.M., Ehsani, R., 2011. Visible-near infrared
spectroscopy for detection of Huanglongbing in citrus orchards. Comput.
Electron. Agric. 77, 127–134.
Tlili, I., Hdider, C., Lenucci, M.S., Riadh, I., Jebari, H., Dalessandro, G., 2011. Bioactive
compounds and antioxidant activities of different watermelon (Citrullus lanatus
(Thunb.) Mansfeld) cultivars as affected by fruit sampling area. J. Food Compos.
Anal. 24, 307–314.
Tu, K., Jancsok, P., Nicolai, B., Baerdemaeker, J..De., 2000. Use of laser-scattering
imaging to study tomato-fruit quality in relation to acoustic and compression
measurements. Int. J. Food Sci. Technol. 35, 503.
Udomkun, P., Nagle, M., Mahayothee, B., Müller, J., 2014. Laser-based imaging
system for non-invasive monitoring of quality changes of papaya during drying.
Food Control 42, 225–233.
Wechter, W.P., Levi, A., Harris, K.R., Davis, A.R., Fei, Z., Katzir, N., Giovannoni, J.J.,
Salman-Minkov, A., Hernandez, A., Thimmapuram, J., Tadmor, Y., Portnoy, V.,
Trebitsh, T., 2008. Gene expression in developing watermelon fruit. BMC
Genomics 9, 275–287.
Zhang, B., Huang, W., Li, J., Zhao, C., Fan, S., Wu, J., Liu, C., 2014. Principles,
developments and applications of computer vision for external quality
inspection of fruits and vegetables: a review. Food Res. Int. 62, 326–343.
Zheng, C., Sun, D.-W., Zheng, L., 2006. Recent developments and applications of
image features for food quality evaluation and inspection – a review. Trends
Food Sci. Technol. 17, 642–655.