Available via license: CC BY-SA 4.0
Content may be subject to copyright.
International Journal of Electrical and Computer Engineering (IJECE)
Vol. 14, No. 3, June 2024, pp. 2803~2813
ISSN: 2088-8708, DOI: 10.11591/ijece.v14i3.pp2803-2813 2803
Journal homepage: http://ijece.iaescore.com
Optimization of CPBIS methods applied on enhanced fibrin
microbeads approach for image segmentation in dynamic
databases
Ramaraj Muniappan1, Thiruvenkadam Thangavel2, Govindaraj Manivasagam2,
Dhendapani Sabareeswaran3, Nainan Thangarasu4, Chembath Jothish5, Bhaarathi Ilango1
1Department of Computer Science, Rathinam College of Arts and Science, Coimbatore, India
2School of Computer Science and IT, Jain University, Bangalore, India
3Department of Computer Science, Government Arts and Science for Women, Tiruppur, India
4Department of Computer Science, Karpagam Academy of Higher Education, Coimbatore, India
5Department of Computer Science, Presidency University, Bengaluru, India
Article Info
ABSTRACT
Article history:
Received Sep 22, 2023
Revised Jan 22, 2024
Accepted Jan 23, 2024
In the empire of image processing and computer vision, the demand
for advanced segmentation techniques has intensified with the
growing complexity of visual data. This study focuses on the
innovative paradigm of fuzzy mountain-based image segmentation, a
method that harnesses the power of fuzzy logic and topographical
inspiration to achieve nuanced and adaptable delineation of image
regions. This research primarily concentrates on determining the age
of tigers, a critical and challenging task in the current scenario. The
primary objectives include the development of a comprehensive
framework for FMBIS and an in-depth investigation into its
adaptability to different image characteristics. This research work
incorporates those domains of image processing and data mining to
predict the age of the tiger using different kinds of color images.
Fuzzy mountain-based pixel segmentation arises from the need to
capture the subtle gradients and uncertainties present in images,
offering a novel approach to achieving high-fidelity segmentations in
diverse and complex scenarios. The proposed methods enable image
enhancement and filtering and are then assessed during process time,
retrieval time, to give a more accurate and reduced error rate for
producing higher results for real-time tiger image database.
Keywords:
Clustering
Data mining
Image classification
Image database
Performance analysis
RGB pixels
Similarity measurements
This is an open access article under the CC BY-SA license.
Corresponding Author:
Ramaraj Muniappan
Department of Computer Science, Rathinam College of Arts and Science
Pollachi Main Rd, Kpm Nagar, Eachanari, Coimbatore, Tamil Nadu 641021, India
Email: ramaraj.phdcs123@gmail.com
1. INTRODUCTION
The current work differs from other approaches in that it involves non-trivial procedures to extract
implicit information from data that has not yet has been discovered and information that may be useful in the
future to understand concepts from a large amount of real-time data that is stored in a database. Data mining
methods include clustering, classification, affiliation, regression, and assessment [1]. There are various useful
data mining techniques like pattern awareness, time tracking, online analytical processing (OLAP),
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 14, No. 3, June 2024: 2803-2813
2804
visualization, and more. Gautam and Singhai [2] interpret it as an interdisciplinary endeavor aimed at better
equipping experts in the domain of artificial intelligence, processing images, acquisition of images,
restoration of images, statistics mining, and gadget gaining knowledge of databases. Kumudham and
Rajendran [3] have referred to the large and detailed image databases created as a result of the rapid
advancements in image capture and storage systems through which users can obtain useful understanding
through image processing. Caponetti et al. [4] have conveyed through the article that image mining has
become complex due to the existence of massive data of different styles of delay stored within the images.
Cong and Hiep [5] has focused on the implementation of support vector machine (SVM) algorithms for
object classification within a robotic arm system, aiming to enhance its cognitive capabilities and broaden its
range of applications.
Wang and Wang [6] has put on record that it facilitated decision-making by developing
relationships, patterns, or clusters. In order to increase the accuracy of the results, this paper concentrates on
investigating the proposed techniques. According to Ramaraj and Niraimathi [7], the enhanced clustering
algorithm can apply the grid-based methods to an image's pixels. Sarrafzadeh and Dehnavi [8] has proposed
mountain clustering rule set that works for grid-based features by using a large number of pixels in the
image. The proposed methods are focused on the mountain clustering (MC) techniques for the most part
based on fuzzy to establish the group amenities by utilizing the function of the peak computationally.
According to Ramaraj and Niraimathi [9], once the hill process is demolished, predicted values for every
important basis that has become adjacent points to the comet's centroids are smaller than the threshold value
for something other than a color.
Al-Ghrairi et al. [10] has sad that, Among the diverse array of applications, object classification
stands out as a crucial task for enabling robots to comprehend and respond to their environment effectively.
Niraimathi [11] has pointed out that, the potency of some of the statistics that may turn out to be potential
clusters and intermediate clusters is reduced to some extent by repeated reductions to the extent that they lose
the possibility of evolving to intermediate clusters. Fahrudin et al. [12] has enhanced k-means clustering
algorithm features of image reconstruction of electrical impedance tomography using simultaneous algebraic
reconstruction techniques, such as no existing deficiency to restrict grid perseverance, and its computational
complexity does not depend on the measurement. Dubey and Mushrif [13] has proposed the most well-
known methods and numerous parameters have been used to improve the results. The final evaluation
methods for the real-time tiger image analysis were applied by the aforementioned methods. As per Said
[14], image mining strategies like shape improvement and partitioning are necessary to mine the tiger's
image. To determine the age of the species, image processing is coupled with statistics methods, in which
data mining executes the circumstances of affairs by reading the statistical document to confirm the age of
the tiger.
2. REVIEW OF LITERATURE
In the field of image segmentation using various techniques, numerous publications have been
published, and many of these references concentrate on unique image segmentation implementations. Murthy
and Hanumanthaiah [15] has reiterated that, one of the natural clustering algorithms is the K-means
algorithm, which uses a variety of procedures to initialize the center but differs in how the center is
determined. Many researchers have taken up the researches on image segmentation and have also produced
better results. Dhanachandra et al. [16] has used the cumulative distribution feature for converting pixel rate
of the images. To make the existing clustering algorithm for the notion or its pixels more accurate and
efficient a new version was created. Dhanachandra and Chanu [17] has portrayed portrays shape
examination as a strategy for settling the essential issue of test acknowledgment; the picture item's
rendering, flip, and rescaling as well as to make sense of the store in connection with finding the item
presented as fringe frames.
Khaleel et al. [18] used K-means cluster analysis to segment images. In order to evaluate the cluster
matrix throughout the segmentation process, this method applied the k-means algorithm. L*a*b area is
superior in assessment with RGB shade space for clustering in precision, bear in mind area. The color based
image segmentation (CBIS) strategy utilized the multi-faceted realities to associate picture pixels into
numerous clusters. Sakarya [19] has proposed a new method for precise whimsy segmentation. With
improved methods and effective techniques based on X-ray imaging, it is possible that this improved
technique could be used to diagnose diseases such as lung cancer and tuberculosis. Sudana et al. [20] by
utilizing the K-means clustering algorithms for the process of grouping and partitioning the image category.
Mazouchi and Milstein [21] has used the above methodologies to carry out the primary color space, while
processing color image segmentation, with the usage of K-means clustering set of rules.
Int J Elec & Comp Eng ISSN: 2088-8708
Optimization of CPBIS methods applied on enhanced fibrin microbeads… (Ramaraj Muniappan)
2805
Gosain and Dahiya [22] for ascertaining starting centroids, the closest feasible information points
were used. The outcomes of the experiments showed that rules can allocate facts into clusters without
requiring much iteration. However, the cluster method has exchanges with the hassle of switching the
magnitude of the preferred by a group of inspiration as a contribution. As per Rani and Bhardwaj [23] has
used the iterative self-organization data analysis techniques (ISODATA) methods' pixel characterization and
the has proposed techniques for image segmentation in the light weight of a pixel by calculating the boundary
assessment that deal with image segmentation in an unmatched way. Tian et al. [24] by utilizing improved
clustering techniques, the illusion set of challenging Balinese consistencies to fulfill the specific progression
in a nearly identical form. Olugbara et al. [25] has told in the article that an enhanced cliff clustering
approach is based primarily on the characteristics of ridge glen functions. The number of cluster centers,
cluster facilities, and statistical patterns associated with each cluster center should be routinely and precisely
gathered by the proposed algorithm.
3. METHOD
3.1. Fuzzy based mountain clustering methods
The grid's choice undoubtedly affects the unique mountain method's clustering effectiveness, with
better performance coming from improved frameworks. However, the cluster increases in cost as a function
of size as the cluster decision is multiplied. Additionally, the initial ridge strategy is computationally
inadequate when applied to excessively dimensional data due to the exponential growth in the numeral of
grid aspects required with record measurements. The algorithm step is given in Figure 1.
Figure 1. Step-by-step process for the proposed clustering algorithm
As illustrated in Figure 1, by using the fuzzy based mountain clustering algorithm, the boosted
classifier rule set can be run and updated. The fuzzy mountain clustering method uses information density
measurements to approximate cluster centers. Mountain clustering method is an independent algorithm that
can also use to locate the initial cluster centers, which may be required when employing fuzzy mountain
clustering (FMC) or other more advanced cluster algorithms. In statistical units with clustering tendencies,
the mountain clustering method is a clustering technique used to approximate the locations of cluster centers
[26].
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 14, No. 3, June 2024: 2803-2813
2806
3.2. Pixel classification function
Utilizing the Mountain method to estimate the cluster and estimate is a specialty of the mountain
function, or degree of density, source. The regulations correspond to the highest possible value that can be
located using a mountain approach [27]. Since, the facilities of the clusters that are received by the peak
characteristic, the mountain approach is responsible for determining the initial estimations of the reference
antecedent parameter and the subsequent fuzzy units of concepts [28].
| U, V|
(1)
(2)
The ith data location and μ for peak radius can be found here. Each data point makes a contribution
to the elevation of the mass process at v. It has an opposite relationship to distance between μ and v, because
μ is an application-specific constant. A record density metrics that can be used to describe the mountain
attributes of μ. The final function's μ, peaks' height and smoothness are determined by constants values are
represented by (1) and (2).
Figure 2 shows that the graphic representation of gimmick to the both distance worth and
development in the iteration process. In Figure 2, each of the iteration process is shown in a different color,
illustrating how they are formed in the different age group of tiger image. Until the various group focuses are
carried out, this method of updating mountain capacities and making decisions for the subsequent cluster
process is to be continuously. The number of clusters is m. By using the parameters
and is and (3),
calculate the prospective value of apiece mountain clustering or segmenting the location point and the other
data points.
Figure 2. Illustrate on proposed clustering methods applied on image classification of the pixels in an image
database
3.3. Pixel transformation function
Pixel transformation techniques can vary widely depending on the specific task and the
characteristics of the image. These transformations play a crucial role in image processing and computer
vision, allowing you to improve image quality, extract information, and create visually appealing results.
(3)
where is the new gray scale image on pixel transformation, the specific mathematical equation for pixel
transformation can be more complex depending on the desired effect and how you choose to blend pixel
values from different clusters.
cluster weight is the degree of membership of the pixel to the cluster.
ClusterPixelValue is the pixel value in the respective cluster to be calculated by the whole
image. TotalWeight is the sum of the cluster weights for the pixel.
Int J Elec & Comp Eng ISSN: 2088-8708
Optimization of CPBIS methods applied on enhanced fibrin microbeads… (Ramaraj Muniappan)
2807
3.4. Color manipulation function
Color manipulation is an important aspect of image editing and computer vision. It allows for the
correction of imperfections, creative expression, and the enhancement of visual content. The choice of color
manipulation technique depends on the specific goals and requirements of the project or image. This can be
done for various purposes, including image enhancement, correction, creative effects, and image analysis.
Color manipulation typically involves altering the color values of pixels, objects, or regions within an image.
Color balance adjustments involve altering the proportions of primary colors (red, green, and blue) to achieve
a more balanced color distribution. It can correct color tints and achieve a more natural look.
(4)
where is an original image of the RGB, for a specific pixel, you have assigned it to clusters with degrees
of membership (). Let's assume have performed fuzzy clustering at the pixel level and have assigned
each pixel to clusters with corresponding degrees of membership (). It will focus on shifting the color of a
single pixel in the red channel (R) while keeping the green and blue channels (G and B) unchanged.
4. RESULTS AND DISCUSSION
The proposed model comes with a tiger image database, which makes using the MATLAB tool
much easier. More than 1,000 different camera trap images, and other source images in a variety of sizes and
different formats, can be found in this database. The different age groups of tigers that have been examples
will all belong to the same class. Each color pixel is divided into color visualizations, which have evolved to
incorporate many colors in a single image, and it splits the individual color pixels into a different window.
4.1. Computational complexity
The quantity of steps in a calculation for the grouping strategy is alluded to as computational
intricacy. The relative effectiveness of different clustering strategies in terms of time complexity is
determined by looking at their computational complexity. i.e., , Compared to other clustering methods,
fuzzy based mountain clustering requires fewer steps to cluster data.
(5)
Accept that N stand for the absolute amount of variety pixels, m for the numeral of groups, and r for
the quantity of cycles on t. The upgraded mountain grouping calculation can diminish the computational
intricacy of each ensuing bunch, bringing about a decrease in handling time, by taking out the bunches from
the dataset and removing the noise ration about the given dataset and predict the age is based on the time of
execution part.
4.2. Find the era calculation of the tiger
The primary difference between the actual value and the standalone value of the ultimate apparatus
that produces the data is the precision of the anticipated technique. The amount of space desirable to separate
the present pixels that have been addressed by the cells in one pixel to another pixel of the inappropriate
frames of (). The unit of measurement for overall segmentation precision can be generated by dividing
these values by the total number of pixels ( . This can be accomplished by determining
( ) the number of pixels in the tiger image database that correspond to the age of the ground.
(6)
For example, d represents the fundamental Euclidean distance, N represents the total number of
pixels in an image, and m represents the number of red, green, blue (RGB) classes of pixels in an image. The
total number of correctly classified pixels in a tiger image is shown here as . Additionally, each color
pixel's threshold value was set to a specific tiger when selecting the tiger image's meticulous age of specific
threshold. With the number of pixels in each row and column, , is represented. To collect the real time
tiger image database for different age group of tigers are stored in an image database. In the real time images
of tiger are more helpful to identify the age of the individual tiger. It based on the skin color; it is supposed to
be RGB color pixels that is used to infer the age of the tiger. Each color pixel is divided into color
visualizations, which have evolved to incorporate many colors in a single image, and it splits the individual
color pixels into a different window.
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 14, No. 3, June 2024: 2803-2813
2808
Table 1 shows how the data is segmented by era. The number of clusters is uniformly assumed to
be 3. The highest precision in the first year is 0.96, while the lowest PP is 0.91. RC distance ranges from 0.93
to 0.97, with 0.97 being the greatest. The highest measured FM is 0.96, while the lowest is 0.94. The table's
similarity measures are slightly elevated when they are compared to one another. The city block has the
lowest F-measure and the Euclidean model has the highest PP (0.96), maximum RC (0.97), and highest FM
is (0.96) respectively. By using similarity-based clustering accuracy measurements, the era of the tiger is
predicted in Table 1. These metrics consist of F-measure, recall, and precision. Additionally, they identify
different parameters such as various parameters were tested with the graph is represented. The experimental
results are shown in Figure 3.
Table 1. Using a sample tiger image and a variety of similarity metrics to determine the tiger's era
Era
SM
PC
RC
FM
1 year
CBD
0.92
0.97
0.95
CCD
0.96
0.93
0.94
ED
0.94
0.95
0.96
MKD
0.91
0.93
0.94
*Note: CBD = City block distance, CCD = Chebychev distance, ED = Euclidean distance, MKD = Minkowski distance.
Figure 3. By using various parameters are tested with 1 year tiger image
The data are categorized by era, as shown in Table 2. The number of clusters is taken to be three
consistently. The highest PP in the second year is 0.96, while the lowest PP is 0.94. The most recall is 0.97,
while the least is 0.94. The lowest FM is 0.945, while the highest is 0.965. The SM in the table is slightly
elated when compared to one another. The most elevated accuracy 0.96 and most noteworthy review 0.97
and the most noteworthy F-measure is 0.965 in ED and the least is found in CC distance.
Illustrate that the given Figure 4 represents the expectation of the era of the tiger by utilizing
proportions of the comparability-based bunching exactness and involving some grouping measurements as
accuracy, review, and FM and figure out the likeness capabilities as a CCD, CC distance, MKD, and ED and
it assists with anticipating the age of the tiger picture. Figure 3 depicts the results of the experiment.
Table 2. Various metrics applied by the sample 2 year tiger image
Era
SM
PP
RM
FM
2 year
CBD
0.94
0.95
0.945
CCD
0.95
0.94
0.945
ED
0.96
0.97
0.965
MKD
0.95
0.96
0.955
*Note: SM = Similarity Measures, PP = Precision, RM = Recall Measure, FM = F-measure.
0.88
0.9
0.92
0.94
0.96
0.98
CBD CCD ED MKD
1 year
0.92
0.96
0.94
0.91
0.97
0.93
0.95
0.93
0.95
0.94
0.96
0.94
Values
Similarity Metrics
one year
PC
RC
FM
Int J Elec & Comp Eng ISSN: 2088-8708
Optimization of CPBIS methods applied on enhanced fibrin microbeads… (Ramaraj Muniappan)
2809
Figure 4. Illustrate on different SM methods applied for the 2-year tiger image
The categorization of the data by year is shown in Table 3. The standard assumption is that there are
three clusters. For 15 years, the precision ranged from 0.91 to 0.96, the highest possible value. The lowest
recall value is 0.97, which falls between 0.90 and 0.97. The lowest and highest F-measures ever recorded are
0.92 and 0.96, respectively. As each performance measures as compared, the similarity measures in the table
feel a little better.
The accuracy of the tiger's age as predicted by similarity-based clustering is shown in Table 3. To
determine the tiger image's age, it makes use of SM methods like CBD, CC distance, MKD, and ED, in
addition to brunch metrics like PP, RCM, FM. Figure 5 depicts the findings of the experiment.
Table 3. Different similarity measures calculated by 15 year tiger image
Age
SMI
PP
RC
FM
15 year
CBD
0.913
0.962
0.936
CCD
0.962
0.954
0.957
ED
0.954
0.976
0.969
MKD
0.947
0.908
0.927
Figure 5. 15-year-old tiger image and various similarity metrics to predict the tiger's era
0.86
0.88
0.9
0.92
0.94
0.96
0.98
CBD CCD ED MKD
15 year
0.91
0.96
0.95
0.94
0.96
0.95
0.97
0.9
0.935
0.955 0.96
0.92
Values
Metrics
PP
RC
FM
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 14, No. 3, June 2024: 2803-2813
2810
In Table 4, the root mean square error (RMSE) value, image retrieval time, and accuracy of both
proposed and existing clustering algorithms are listed along with their respective accuracies. The clustering
results are displayed in the form of plots when the presented algorithms are implemented, and the final
product is significantly more useful and effective. The filter bank multicarrier (FBMC) accuracy of the
results from the suggested methods is the highest. Figure 6 shows the overall performance of the tiger image
database in terms of accuracy, RMSE, time, and image retrieval, as well as comparisons to suggested and
existing methods.
Table 4. Overall measurements
Algorithm Measurements
AR
ERMSE
TRP
IMR
EM
94.5
EM
0.667
EM
1.877
EM
1.36
PM
97.6
PM
0.333
PM
1.032
PM
0.865
*Note: AR = Accuracy, ERMSE = Enhanced root mean square error, TRP = Time per second, IMR = Image retrieval,
EM = Existing methods, PM = Proposed methods.
Figure 6. Illustrate on overall performance of proposed and existing methods
5. CONCLUSION
The ultimate goal of this paper is to find the age of the tiger, which stands primarily found on tiger
image databases. This research article is specifically absorbed on the methods anticipated that have
accumulated the more than 1,000+ real-time tiger images that have been collected within the flora and fauna
woodland, and the varied photographic styles of different grown-up tigers have been analyzed. Because of
the various color combinations of the tiger pigmentation, and patterns present in the tiger images representing
different chronological ages, it is possible to determine whether the clustering was achievable and whether it
was possible to determine the exact age of the tiger. The fuzzy clustering methods presented in the future
sections are mostly used in the age estimation of tigers based on the color of the tiger image. Each image is
grouped using the distinction in age and color provided by clustering. Fuzzy clustering methods discussed in
the prior section can be utilized to estimate a tiger's age based on its color in the tiger image database. The
age estimation of tigers is based on the color of their images and frequently depends on significantly used
fuzzy clustering models that will be explored in the sections of this article immediately proceed. The original
image demonstrated that the suggested strategy is effective in terms of accuracy and execution time when
compared to its most current effectiveness in an enhanced presentation of the new scientific approach. Image
analyses are used to show how well the suggested clustering algorithm performs in terms of accuracy and
computation time so that it may be contrasted with the effective performance of the suggested strategies. The
clustering result is extremely potent, green, and mentioned in the outcomes phase.
0
10
20
30
40
50
60
70
80
90
100 94.5 97.6
0.667
0.333
1.877
1.032
1.36
0.865
Values
Paramenters Series1
Int J Elec & Comp Eng ISSN: 2088-8708
Optimization of CPBIS methods applied on enhanced fibrin microbeads… (Ramaraj Muniappan)
2811
REFERENCES
[1] S. Yang, P. Li, H. X. Wen, Y. Xie, and Z. He, “K-hyperline clustering-based color image segmentation robust to illumination
changes,” Symmetry, vol. 10, no. 11, pp. 1–11, 2018, doi: 10.3390/sym10110610.
[2] K. Gautam and R. Singhai, “Color image segmentation using particle swarm optimization in lab color space,” International
Journal of Engineering Development and Research, vol. 6, no. 1, pp. 373–377, 2018.
[3] R. Kumudham and V. Rajendran, “Implementation of various segmentation algorithms on side scan sonar images and analysing
its performance,” ARPN Journal of Engineering and Applied Sciences, vol. 12, no. 8, pp. 2396–2400, 2017.
[4] L. Caponetti, G. Castellano, and V. Corsini, “MR brain image segmentation: A framework to compare different clustering
techniques,” Information (Switzerland), vol. 8, no. 4, 2017, doi: 10.3390/info8040138.
[5] V. D. Cong and T. T. Hiep, “Support vector machine -based object classification for robot arm system,” International Journal
of Electrical and Computer Engineering (IJECE), vol. 13, no. 5, pp. 5047–5053, 2023, doi: 10.11591/ijece.v13i5.pp5047-
5053.
[6] Y. Wang and C. Wang, “High resolution remote sensing image segmentation based on multi-features fusion,” Engineering
Review, vol. 37, no. 3, pp. 289–297, 2017.
[7] M. Ramajac and S. Niraimathi, “Design and implementations of color pixel based image segmentation using enhanced data
clustering algorithms to applying on tiger image dataset,” International Journal of Pure and Applied Mathematics, vol. 119,
no. 18, pp. 2719–2739, 2018.
[8] O. Sarrafzadeh and A. Dehnavi, “Nucleus and cytoplasm segmentation in microscopic images using K-means clustering and
region growing,” Advanced Biomedical Research, vol. 4, no. 1, p. 174, 2015, doi: 10.4103/2277-9175.163998.
[9] M. Ramaraj and S. Niraimathi, “Comparative analysis of proposed methods for analysing color pixel based image segmentation
using tiger image dataset,” International Journal of Research, vol. 7, no. 9, pp. 245–252, 2018.
[10] A. H. Thary Al-Ghrairi, N. H. F. Al-Anbari, and E. Z. Sameen, “Realtime human face tracking and recognition system on
uncontrolled environment,” International Journal of Electrical and Computer Engineering, vol. 13, no. 4, pp. 4246–4255, 2023,
doi: 10.11591/ijece.v13i4.pp4246-4255.
[11] S. Niraimathi, “Enhanced image segmentation to infer the age of the tiger using fuzzy modified K -means clustering algorithm
(FMKMCA),” Research Journal of Pharmaceutical, Biological and Chemical Sciences, vol. 10, no. 2, p. 198, 2019.
[12] A. E. Fahrudin, E. Endarko, K. Ain, and A. Rubiyanto, “Enhanced image reconstruction of electrical impedance tomography
using simultaneous algebraic reconstruction technique and K-means clustering,” International Journal of Electrical and
Computer Engineering (IJECE), vol. 13, no. 4, pp. 3987-3997, Aug. 2023, doi: 10.11591/ijece.v13i4.pp3987-3997.
[13] Y. K. Dubey and M. M. Mushrif, “FCM clustering algorithms for segmentation of brain MR images,” Advances in Fuzzy
Systems, vol. 2016, 2016, doi: 10.1155/2016/3406406.
[14] N. S. M. Said, H. Madzin, S. K. Ali, and N. S. Beng, “Comparison of color-based feature extraction methods in banana leaf
diseases classification using SVM and K-NN,” Indonesian Journal of Electrical Engineering and Computer Science (IJEECS),
vol. 24, no. 3, pp. 1523-1533, Dec. 2021, doi: 10.11591/ijeecs.v24.i3.pp1523-1533.
[15] P. Narasimha Murthy and S. K. Yeliyur Hanumanthaiah, “A simplified and novel technique to retrieve color images from hand -
drawn sketch by human,” International Journal of Electrical and Computer Engineering (IJECE), vol. 12, no. 6, pp. 6140-6148,
Dec. 2022, doi: 10.11591/ijece.v12i6.pp6140-6148.
[16] N. Dhanachandra, K. Manglem, and Y. J. Chanu, “Image segmentation using K-means clustering algorithm and subtractive
clustering algorithm,” Procedia Computer Science, vol. 54, pp. 764–771, 2015, doi: 10.1016/j.procs.2015.06.090.
[17] N. Dhanachandra and Y. J. Chanu, “Image segmentation method using K-means clustering algorithm for color image,” Advanced
Research in Electrical and Electronic Engineering, vol. 2, no. 11, pp. 68-72, 2015.
[18] A. D. Khaleel et al., “Pattern reconfigurable dielectric resonator antenna using capacitor loading for internet of things
applications,” International Journal of Electrical and Computer Engineering (IJECE), vol. 13, no. 4, pp. 4145–4151, 2023, doi:
10.11591/ijece.v13i4.pp4145-4151.
[19] O. Sakarya, “Applying fuzzy clustering method to color image segmentation,” Proceedings of the 2015 Federated Conference on
Computer Science and Information Systems, FedCSIS 2015, vol. 5, pp. 1049–1054, 2015, doi: 10.15439/2015F222.
[20] O. Sudana, I. K. G. D. Putra, M. Sudarma, R. S. Hartati, and A. Wirdiani, “Image clustering of complex balinese character wit h
DBSCAN algorithm,” Journal of Engineering Technology, vol. 6, no. 1, pp. 548–558, 2018.
[21] A. Mazouchi and J. N. Milstein, “Fast optimized cluster algorithm for localizations (FOCAL): a spatial cluster analysis for super-
resolved microscopy,” Bioinformatics, vol. 32, no. 5, pp. 747–754, 2016, doi: 10.1093/bioinformatics/btv630.
[22] A. Gosain and S. Dahiya, “Performance analysis of various fuzzy clustering algorithms: a review,” Procedia Computer Science,
vol. 79, pp. 100–111, 2016, doi: 10.1016/j.procs.2016.03.014.
[23] P. Rani and R. Bhardwaj, “An approach of colour based image segmentation technique for differentiate objects using MATLAB
simulation,” International Journal of Advanced Research in Computer and Communication Engineering, vol. 5, no. 7,
pp. 553–556, 2016.
[24] L. Tian, L. Han, and J. Yue, “Research on image segmentation based on clustering algorithm,” International Journal of Signal
Processing, Image Processing and Pattern Recognition, vol. 9, no. 2, pp. 1–12, 2016, doi: 10.14257/ijsip.2016.9.2.01.
[25] O. O. Olugbara, E. Adetiba, and S. A. Oyewole, “Pixel intensity clustering algorithm for multilevel image segmentation,”
Mathematical Problems in Engineering, vol. 2015, 2015, doi: 10.1155/2015/649802.
[26] Z. Chen, Z. Qi, F. Meng, L. Cui, and Y. Shi, “Image segmentation via improving clustering algorithms with density and
distance,” Procedia Computer Science, vol. 55, pp. 1015–1022, 2015, doi: 10.1016/j.procs.2015.07.096.
[27] S. Dasari, A. R. M. Reddy, and B. E. Reddy, “Two way clustering based on minimum spanning tree and dbscan algorithm for
image mining,” International Journal of Scientific Research Engineering & Technology (IJSRET), vol. 6, no. 12, pp. 1140–1146,
2017.
[28] S. K. Dubey and S. Vijay, “A review of image segmentation using clustering nethods,” International Journal of Applied
Engineering Research, vol. 13, no. 5, pp. 2484–2489, 2018.
ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 14, No. 3, June 2024: 2803-2813
2812
BIOGRAPHIES OF AUTHORS
Ramaraj Muniappan is working as an assistant professor in the Department of
Computer Science at Rathinam College of Arts and Science, Coimbatore. He holds a Ph.D.
degree in computer science at Bharathiar University in the year of 2020 with specialization in
data mining with image process and also fuzzy logic in the image analysis. His research areas
are data mining, image processing, fuzzy logic, pattern recognition and deep learning concept.
He has published more research article in the reputed various national and international
journals and also filed the patents in the same field. He has a reviewer of many international
journals including with IEEE, ASTESJ, and JERS. He can be contacted at email:
ramaraj.phdcs@gmail.com, ramaraj.phdcs123@gmail.com.
Thiruvenkadam Thangavel associate professor, School of CS & IT, JAIN
(Deemed-to-be University), Bengaluru holds Ph.D., (computer science) has more than 16 years
of experience in academics includes 4 years of international teaching exposure. His areas of
interest include computer networks, operating systems, network administration, Linux
administration, virtualization, cloud computing, cloud security and software engineering. He
has published 2 books and completed 2 funded projects and published 16 research articles in
well-known journals, indexed in the Scopus, UGC Care List. He has attended more than
25 international and national conferences and workshops He has completed an associate
level global certification in AWS Solutions Architect. He can be contacted at email:
mailone.thiru@gmail.com.
Govindaraj Manivasagam is an associate professor at Jain University, has a
significant experience of working for 17 years in academics and research space. His
specialized delivery expertise in her areas of interests such as software testing, data science.
He has over 14 research publications in well-known journals, indexed in the Scopus, UGC
Care List, and has a few patents to his credentials. He has attended more than 5 international
and 20 national conferences and organized a number of seminars as well as numerous FDPs
and workshops. He is a strong education professional with numerous online certification
courses from various platforms. He can be contacted at email: mani.mca.g@gmail.com.
Dhendapani Sabareeswaran is working as an assistant professor in the
Department of Computer Science at Government Arts and Science for women, Tiruppur. He
holds a Ph.D. degree in computer science at Karpagam Academy of Higher Education in the
year of 2020 with specialization in data mining. His research areas are data mining, image
processing, fuzzy logic, pattern recognition. He has published more research article in the
reputed various national and international journals and also filed the patents in the same field.
He can be contacted at email: sabaredhandapani@gmail.com.
Nainan Thangarasu is currently working as an assistant professor in the
Department of Computer Science, at Karpagam Academy of Higher Education, Coimbatore.
He is greatly fascinated with the advanced computing technology and research programs is
cluster computing, cryptography and network security, cloud computing, artificial intelligent
system, information security in large database and data mining as well as the strong teaching
experience. His doctoral dissertation also focuses on advanced security systems with cloud
computing, and he has published more than 13 publications in reputed journals, which he find
would be a great addition to the success of your teaching and research department. He can be
contacted at email: drthangarasu.n@kahedu.edu.in.
Int J Elec & Comp Eng ISSN: 2088-8708
Optimization of CPBIS methods applied on enhanced fibrin microbeads… (Ramaraj Muniappan)
2813
Chembath Jothish completed his PhD from Karpagam deemed to be University,
Coimbatore, Tamilnadu in the year 2020, under the title “Next web page prediction using
enhanced preprocessing and ensemble clustering based hybrid Markov model using web log
data”. More than 15 research journals have been published in reputed journals which are either
Scopus indexed, Web of Science and International Journals of repute, which would help the
Internet mechanism to predict the user’s intention and interest, when Internet is browsed using
mathematical model algorithms after cleaning the internet data, thereby giving accurate
predictions using mathematical models. Author also has attended 3 international conferences
held in the Sultanate of Oman, Malaysia and Cochin. Currently author is working in Presidency
University, Bengaluru. The teaching and research experience is around two decades starting from
2003 to toll date. He can be contacted at email: jothishchembath12@gmail.com.
Bhaarathi Ilango is working as an assistant professor in the Department of
Computer Science at Rathinam College of Arts and Science, Coimbatore. He holds a NET.
Qualification in computer science at NTA in the year of 2023. with specialization in data
mining with image process and also fuzzy logic in the image analysis. His research areas are
data mining, in the various fields. He has published more research article in the reputed
various national and international journals and also filed the patents in the same field. He can
be contacted at email: bhaarathi.cs@rathinam.in.
