Global Journal of Computer Science and Technology: F

Graphics & vision

Volume 17 Issue 2 Version 1.0 Year 2017
Type: Double Blind Peer Reviewed International Research Journal
Publisher: Global Journals Inc. (USA)
Online ISSN: 0975-4172 & Print ISSN: 0975-4350

Color Image Segmentation using Automated K-Means Clustering

with RGB and HSV Color Spaces
By Md. Rakib Hassan, Romana Rahman Ema & Tajul Islam
North Western University

Abstract- Segmentation implies the division of an image into different objects or connected regions
that do not overlap. Though, extensive research has been done in creating many different
approaches and algorithms for image segmentation, however, it is still not very clear to assess
whether one algorithm produces more accurate segmentations than another, whether it be for a
particular image or set of images, or more generally, for a whole class of images [7]. A reliable and
accurate segmentation of an image is, in general, very difficult to achieve by purely automatic means.
Present researches on image segmentation using clustering algorithms reveals that K-means
clustering algorithm so far produces best results but some improvements can be made to improve
the results. The biggest disadvantage of our heavy usage of k-means clustering, is that it means we
would have to think of a k each time, which really doesn’t make too much sense because we would
like to algorithm to solve this on his own. Therefore we tried to find the K automatically and so create
segmentation without any human giving “hints” to the algorithm.
Keywords: automated k-means, clustering, RGB, HSV, segmentation, color space, cluster, image
processing, color image, K-means clustering.
GJCST-F Classification: I.2.10, I.3.3

ColorImageSegmentationusingAutomatedKMeansClusteringwithRGBandHSVColorSpaces
Strictly as per the compliance and regulations of:

© 2017. Md. Rakib Hassan, Romana Rahman Ema & Tajul Islam. This is a research/review paper, distributed under the terms of the
Creative Commons Attribution-Noncommercial 3.0 Unported License http://creativecommons.org/licenses/by-nc/3.0/), permitting all
non-commercial use, distribution, and reproduction inany medium, provided the original work is properly cited.
 Color Image Segmentation using Automated
K-Means Clustering with RGB and HSV
Color Spaces
Md. Rakib Hassan α, Romana Rahman Ema σ & Tajul Islam ρ

2017
Abstract- Segmentation implies the division of an image into objectives like reduce overall deviation, maximize
different objects or connected regions that do not overlap. property, reduce the options or reduce the error rate of

Year
Though, extensive research has been done in creating many the classifier. In image processing, it is still not very clear
different approaches and algorithms for image segmentation, to measure whether one algorithm express more
however, it is still not very clear to assess whether one
accurate segmentation than another, whether it be for 33
algorithm produces more accurate segmentations than
another, whether it be for a particular image or set of images,
related image, particular image or set of images, or

Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I
or more generally, for a whole class of images [7]. A reliable more usually, for a whole class of each images [4].
and accurate segmentation of an image is, in general, very
difficult to achieve by purely automatic means. Present
researches on image segmentation using clustering
algorithms reveals that K-means clustering algorithm so far
produces best results but some improvements can be made
to improve the results. The biggest disadvantage of our heavy
usage of k-means clustering, is that it means we would have to
think of a k each time, which really doesn’t make too much Figure 1: Segmentation Technique with automated “K”
sense because we would like to algorithm to solve this on his
own. Therefore we tried to find the K automatically and so The modern researches on digital image
create segmentation without any human giving “hints” to the segmentation generally use K-means clustering
algorithm. So we tried to make the process automatic. In this algorithm which doesn’t express best results but after
paper, the combined segmentation of RGB and HSV color some elevation of K-means algorithm, it could express
spaces give more accurate segmentation result compared to the results better. In this paper, the combined
segmentation of single color space. For keeping the k
segmentation of RGB and HSV color spaces give more
parameter as small as possible, we had to keep different
intensity levels of the same color on the same segment to
accurate segmentation result compared to
estimate the right k automatically for the algorithm. segmentation of single color space our main objectives
Keywords: automated k-means, clustering, RGB, HSV, while segmenting images of different kind like using the
segmentation, color space, cluster, image processing, RGB color space where we wanted to get rid of different
color image, K-means clustering. intensity level of the same color shade which is not
satisfied. Hence, we thought of another representation
I. Introduction to color images the HSV color space which will ease our
solution to segmentation

W
ith the impetuous improvement in the digital
technology, digital image play very significant II. Related Work
role in modern era with its rapidly uses at
medical sector and the visualization sector. Digital Gudrun J. Klinker, Steven A. Shafer and Takeo
image can be described as a large number array of Kanade observed Physical Approach to Color Image
discrete dots where each dot has a brightness level understanding with additional example through
associated with it. opposed to k means and Anil Z Chitade evaluated the
These dots are simply represents as pixels or Color based image segmentation using kmeans
picture elements. Image segmentation is defined as; clustering. R. C. Gonzalez and R. E. Woods described
“the search for homogeneous regions in an image and Digital Image Processing where YZang presented a
later the classification of these images”. Segmentation review on image segmentation techniques with remote
signifies the partition of an image into different objects sensing perspective.
or connected regions that do not overlap. Highly use of Pena M, Barbakh W, Fyfe C discussed the
Real world image segmentation issues have multiple Principal Manifolds for Data Visualization and Dimension
Reduction. McAndrew A analyzing Digital Image
Author α σ ρ: Dept. of Computer Science and Engineering (CSE), North Processing with Matlab Notes for SCM2511 Image
Western University, Khulna, Bangladesh. e-mails: [email protected], Processing. Marques, Blanca NP, Pina P works with
[email protected], [email protected]

© 2017 Global Journals Inc. (US)

 Color Image Segmentation using Automated K-Means Clustering with RGB and HSV Color Spaces

Pattern Recognition and Image Analysis MacQueen J

proposed some methods for classification and analysis
of multivariate observations.
Y. Li and Y. Shen done a great job about
Robust image segmentation algorithm using fuzzy
clustering based on kernel-induced distance measure
whereas D. E. Ilea and P. F. Whelan gave a complete
overview about Color image segmentation using a
spatial k-means clustering algorithm.
N. R. Pal and S. K. Pal gave A review on image
segmentation techniques and Pattern Recognition in Figure 2: Partioning and Hierarchical Clustering
2017

their paper and H. C. Chen and S. J. Wang works with

Visible color difference-based quantitative evaluation of b) K-Means Clustering
Year

color segmentation, The k-means algorithm works like to cluster "n"

number of objects based on characteristic into k
34 III. Description of Image Segmentation partitions, where k < n. Simply speaking it is an
Features algorithm to group or to classify objects based on their
Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I

corresponding features into K number of group. Here K

a) Clustering is always positive integer number [16]. In general, the
Clustering (also called unsupervised learning) is group is created based on minimizing the sum of
the method of partitioning a data-set into various distances between the corresponding clusters centroid
groups. Sometimes the members of every group are as and data [5] [11].
similar (close) as probable to one another, and
sometimes different groups are as dissimilar (far) as
possible from one another [9]. Clustering can discover
previously unperceived relationships in a data set. There
are various applications for cluster analysis. Cluster
analysis can be used to discover and distinguish
customer segments for marketing purposes. In Biology,
it uses at alignment of plants and animals given their
features. Not only that it can also be used in pattern
recognition spatial data analysis, image processing etc.
[20] In other words, clustering means collection of data
or distribution a large data set into smaller data sets of
some similarity [5]. An important appearance in
clustering is how to designate the similarity between
individual objects, Clusters can be made from objects
with the possibility of high similarity within clusters and
low similarity between clusters [7]. Commonly, to
measure similarity or dissimilarity between objects there
are many popular methods, such as Euclidean, Figure 3: Procedures of K-Means Clustering Algorithm
Manhattan and Minkowski are used. A distance function
executes a lower value for similar objects that are more c) Basic Steps of K -Means Clustering Algorithm
similar to one another. Requirements of good cluster: Given n objects, choose k number of clusters to
be determined. Choose k objects randomly as the initial
• The ability to discover some or all of the hidden
cluster. Assign each object to their closest cluster
clusters also handle high dimensionality.
center. Update the center of each cluster (Calculate
• Within-cluster similarity and between-cluster
mean points) until no changes on cluster centers (i.e.
dissimilarity and Scalable, Interpretable and usable
Centroids do not change location any more). A number
• Ability to deal with various types of attributes, can
of other strategies stands for verifying K, including
deal with noise and outliers.
cross-verify, information criteria, the information the
Hierarchical Clustering is a set of nested silhouette procedure, theoretic jump procedure and also
clusters organized as a hierarchical tree and on the the G-means algorithm. In addition, monitoring the
other hand Partitioning Clustering is a division data classification of data points across various groups
objects into non-overlapping subsets (clusters) such provides acumen into how the algorithm is partitioning
that each data object is in exactly one subset [4]. the data for each K.

© 20 17 Global Journa ls Inc. (US)

 Color Image Segmentation using Automated K-Means Clustering with RGB and HSV Color Spaces

d) Choosing “K” of K -Means Clustering Algorithm

The algorithm described over discovers the
clusters and data set labels for an individual pre-chosen
K. To discover the number of clusters in the data, the
user needs to operate the K-means clustering algorithm
for a range of K values and compare the outcomes. In
general, there is no procedure for recognizing exact
value of K [18].

2017
Figure 5: HSV Color Space

Year
V represents Value which percentage is
between 0 to 100. This percentage range can be
concern as the amount of light expressing a color. For 35
example, the value is high and when the hue is red, the

Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I
color seems bright. On the other hand, it looks dark
when the value is low. Following figures show the HSV
converted image of the original image [5]:
Figure 4: Elbow Point Example
One of the metrics that is generally used to
compare outcomes across several values of K is the
mean distance between data points and their cluster
centroid. Since increasing the number of clusters will
always maximize this metric and minimize the distance
to data points to the ultimate of reaching zero when K is
the same as the number of data points. Thus, this
metrics cannot be used as the absolute target. Figure 6: Sample to HSV Converted Result

e) HSV Color Space f) RGB Color Space

HSV color space will be more compatible for A color image f∈RN×3 is made of three
conduct with segmentation of rough color images. We independent images, one for each channel red, green
can describe HSV color space with the help of a hex and blue which is called RGB color space [11]. Here
cone with three extensions where the middle vertical axis Size N= n×n of the image; where n = 256; N = n*n.
describes the intensity [level 18]. Here H represents RGB is the most widely used color space, that each
Hue. Basically Hue is an angle, which range is [0,2 π] color image is combination of three different images,
comparative to angle 0 at red axis, 2 π/3 at green axis, those are Red image, Blue image and black image [12].
4π/3 at blue axis and red again at 2π. S represents
Saturation, which represents how authentic the hue is
with respect to a white section. This can be concern of
as the depth or integrity of color and is measured as a
long range distance from the middle axis with properties
between 0 at the center to 1 at the outer surface [13].
While, for a given intensity and hue, if the saturation is
qualified from 0 to 1, the comprehend color switch from
a shade of gray to the most authentic form of the color
represented by its hue. Diagrammatic view is given
below [5]: Figure 7: RGB Color Space
RGB color space is also called additive color
space, which can be described well based on the RGB
color model [10]. Three chromaticity's is represented by
a particular RGB color space which includes the red,
green, and blue additive primaries, and also generate
any chromaticity with a triangle which represented by

© 2017 Global Journals Inc. (US)

 Color Image Segmentation using Automated K-Means Clustering with RGB and HSV Color Spaces

those fundamental colors. A gamma correction curve

and a white point chromaticity also requires for the total
specification of an RGB color space [12].
2017
Year

Figure 8: Sample to RGB Converted Result

36
Three colored lights for red, green, and blue are
Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I

produced and described from An RGB color space, for

example, shining three lights together onto a white wall
in a dark room: one red light, one green light, and one
blue light, each with dimmers. If only the red light is on,
the wall will be red. If only the green light is on, the wall
will look green. If the red and green lights are on
together, the wall will look yellow. Dim the red light and
the wall will become more of a yellow-green. Dim the
green light instead, and the wall will become more
orange. Typical RGB input devices are color TV and
video cameras, image scanners, video games, and
digital cameras.
Figure 9: Proposed Technique of RGB Color
IV. Methodology Segmentation
To generate the best practical segmentation, we
To reduce the K parameter, we start with a high
intend to use images with known characteristics, where
value of K, and iteration ally reducing it, until we go an
most popular use about the segmentation is color. Our
indicated condition, which normally has to do with the
aim is to find out an amusing way to segment the image
minimum distance between the centers of the
by the colors with the k means clustering algorithm,
segmentation resulted by the k-means algorithm. For
which includes mapping the image pixels to the RGB
color space and HSV color space (i.e. if pixel 4, 5 holds calculate the sample values of RGB and HSV we used
the values red: 50, green: 30, blue: 20, we map it to the Matlab. Of the colors we wanted to separate, we tried to
point (50, 30, 20) on a 3d space representing the RGB take the algorithm to a point where it wouldn’t
and HSV values). Our proposed function works like to accumulate those colors under the corresponding
adjust the input to the format of k means procedure for segment. If we consider the minimum distance between
expects to get and ran a few tests. Our motivation was the colors then we can use a condition that will force the
to keep the k parameter as small as possible to estimate k to be small enough that the various segments
the right k automatically for the algorithm. For that, we generated will be at least as far from one another, as the
had to keep different intensity levels of the same color various colors of the game play world, we won’t get a
on the same segment; for example trying to keep light segment with colors “close to one another”. We
green and dark green in the same segment. Our basic confined heavily on our basic supposition, associated
supposition for this part of the work, was that whether with hope that various intensity of the same color may
given a data set of n clusters, when we operate K be closer to one another, after decreasing K value and
means algorithm with parameter k=m (Where m>n). increases the minimum distance our proposed
The centers which is created by the algorithm would be algorithm is ends. Basically various “shades” of the
closer to each other than if we ran the algorithm with the same color are multiplication by a constant of a basic
parameter k=n (the actual number of clusters). So the vector in the RGB space. If in the data set there are
minimum distance between the cluster centers would several clusters, each one representing several intensity
decrease. level of the color.

© 20 17 Global Journa ls Inc. (US)

 Color Image Segmentation using Automated K-Means Clustering with RGB and HSV Color Spaces

condition based on our knowledge of the game play

images, like the colors existing in the world.

2017
Year
Figure 11: Sample Image from Fruit-Ninja Game
On fruit ninja game play image we sampled the 37
different colors from all the individual fruits in the game,

Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I
and using the minimum distance between the fruits as a
starting point for a terminating condition. We noticed
with this result, the perceive areas to be segment and
assigned to in the result. The background got divided to
two segments. Here someone can argue that if we
segment the image based only on colors, but we
wanted it to be a good pre-processing for a later by
object recognition algorithm. But as we started
implementing this method, we realized that there is
much simpler solution.
Figure 10: Proposed Technique of HSV Color
Segmentation
Our main goal is to get rid of various intensity
level of the same color hue. Hence, the illustration to
color images the HSV color space which will comfort our
solution to segmentation [8]. The HSV color space has
several parameters for the HUE, SATURATION and
VALUE (intensity) of the pixel, and so we get instinctive
access to the actual hue of the pixel, without any
concerns about several intensity level.

V. Performance Results Figure 12: HSV Result with Automated K after Color
a) Performance Matrices Segmentation
We discovered that even though the K-Means Hence, we thought of another representation to
algorithm does a great job, but when we tried to make color images the HSV color space which will ease our
the process automatic with the right automatic K, even solution to segmentation. The HSV color space has
though we were in a restricted world of images, the different parameters for the HUE, the SATURATION and
process wasn’t easy, and obviously our results are not the VALUE (intensity) of the pixel. Again, the fully
perfect, but not too far away. On images of unknown automatic algorithm isn’t perfect, the usage of the
source, the problem becomes much harder. So we minimum distance condition isn’t enough to accurately
proposed an algorithm to minimize the K parameter by estimate the K parameter but we do ideas on how to
itself, at the beginning of time it starts with high value of improve it. In the restricted world of game play images,
K, and then iteration ally reducing it, until we meet a HSV based segmentation did much better than the RGB
certain condition, which generally has to do with the version. The fact that we can ignore the pixels intensity
minimum distance between the centers of the level opened up a way to clearly segment the picture
segmentation. Now, the problem is to choose a based on hue only, which, in this scenario specifically is
condition which will make the algorithm stop and very good.
calculate the “right” k. We can easily fix this certain

© 2017 Global Journals Inc. (US)

 Color Image Segmentation using Automated K-Means Clustering with RGB and HSV Color Spaces

From the figure number 12 and 13 we get a

clear view that our proposed K-means clustering
algorithm give better output on Game play images and
also can segment properly without giving any human
hints of find out the number of ”K."
c) Working Results when RGB is better than HSV
Segmentation by colors generally show a very
perfect sense of the picture, and can be a decent
preprocessing for stages like object visualization. The
Figure 13: RGB Result with Automated K after Color role of segmentation is authentic in most tasks requiring
Segmentation image analysis. The success or failure of the process is
2017

No representation is better than the other often a direct result of the success or failure of
Year

though we wanted to emphasize this fact, so we provide segmentation. However, an authentic and exact
examples from real world images; there HSV color segmentation of an image is, in general, very difficult to
acquire by purely automatic means. Considering the
38 space to eliminate different intensities of the same hue
isn’t performing too well, as opposed to RGB. pros and cons of the main algorithm used (K-Means
Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I

Clustering) and the fact that segmentation can be

b) Working Results when HSV is better than RGB understand differently by different people, we set our
The major step we did was the HSV color space minds to several goals. When we worked with Real
to get a clearer segmentation view to us, for better out world Images our RGB results were so much better than
come in the future we like to add the K recognition HSV Color Space method. We got proper
feature which still requires work to be fully functional. segmentations. Here some results are given below:
Here is several samples of the results we achieved, with
some annotations:

Figure 14: HSV Result on Fish image after Color

Segmentation
We know about the importance of having
different representations of the same object, and how Figure 16: RGB Result on Flower image after Color
each one opens up a world of tools to use on it, lets you Segmentation
get a part of the information easily, while hiding some of
the redundant information. For the segmentation we have proposed the
idea of k means clustering algorithm which involves
mapping the image pixel to the RGB color space and
HSV color space. We proposed some function to set the
input to the format the k-means iteration expects, and
ran few tests where the outputs were quite impressive
[8]. For keeping the k parameter as small as possible,
to estimate the right k automatically for the algorithm, we
had to keep various intensity levels of the color on the
same segment.

Figure 15: HSV Result on Bird image after Color

Segmentation

© 20 17 Global Journa ls Inc. (US)

 Color Image Segmentation using Automated K-Means Clustering with RGB and HSV Color Spaces

On RGB color space, the optimal K value and

the Automated K value was not so close. So on game
play images, our proposed automated K-Means
algorithm could not provide impressive result; On the
other hand when we worked with automated K-Means
on HSV Color space, we got the K value so much closer
with optimal K-Means value. So most of the times it
gave us impressive result and also achieved our goal.
But it is difficult to say HSV result is all time better than
RGB result with our proposed automated K-Means
algorithm. Although K-means algorithm does

2017
comparatively better task but the process commits little
bit ambiguous while we have try to choose the actual k.

Year
We efficient about various ways to select the right "K" for
the algorithm, like G-Means clustering, which uses
statistics and hypotheses about the data to calculate the 39
Figure 17: RGB Result on Fruit image after Color
K, and X-means algorithm which didn’t supposed to
Segmentation

Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I
give a result that would glut [2]. So we tried to make the
We know that different “shades” of the same method instinctive, even though we were in a strict world
color (different intensities) are basically multiplication by of images, the work is not easy and accurate but not
a constant of a basic vector in the RGB space. The most impossible by any means. When worked with RGB color
disadvantage of our excessive usage of k-means space, our proposed technique is not gave us suitable
clustering is that it means we would have to think of a k result. The optimal K value and the automated K value is
every time, which really not the perfect way because we quite difference.
would like to proposed the algorithm to solve this by
own. Therefore we attempted to get the K automatically
and so express segmentation without any user giving
“hints” to the algorithm. As we explained before, for
these scenarios (Game play images), RGB isn`t
Performing exceptionally well.
d) Comparison between optimal K-Means and
Automated K-Means Algorithom
For the scenario (Game play images), RGB
isn`t Performing well. If the color set is known then for
the same color hue we don’t need to give different
Segments to different intensity levels, K-Means
algorithm on the HSV color space does more impressive
job. From below example we can easily compare the K
value between Optimal K-Means and our Automated K-
Means at RGB and HSV Color space. Figure 19: Optimal and Automated K value on HSV
Work with normal K-means algorithm has quite
few problems regarding the color segmentation of
images that it we would have to think of "K" each time for
segmentation result, which is one of the biggest
disadvantage of k-means clustering algorithm. Basically
it doesn’t make too much intuition because we would
like to do algorithm to execute this by own. When we
worked with HSV Color space, the optimal K vale and
the automated K value is quite same which assure us
that our automated K-means algorithm performed better
at HSV color space.

VI. Conclusion and Future Work

We discovered that even though the K-Means
Figure 18: Optimal and Automated K value on RGB algorithm does a great job, the fact that we have to

© 2017 Global Journals Inc. (US)

 Color Image Segmentation using Automated K-Means Clustering with RGB and HSV Color Spaces

choose a k makes using it a bit problematic. When we Applications (0975 – 8887) Volume 41– No.20,
tried to make the process automatic but the job wasn’t March 2012.
easy, and obviously our results are not perfect, but not 6. Dibya Jyoti Bora1, Anil Kumar Gupta2, Fayaz
too far away. The characteristics which we use for our Ahmad Khan3 "Comparing the Performance of
working procedure is color: As most of us know, the L*A*B* and HSV Color Spaces with Respect to
different figures in the game are fruits, which usually Color Image Segmentation" , International Journal of
have very different colors, and so segmentation by Emerging Technology and Advanced Engineering
colors usually gives a very correct feel of objects, and Website: www.ijetae.com (ISSN 2250-2459, ISO
can be a good preprocessing for stages like object 9001:2008 Certified Journal, Volume 5, Issue 2,
recognition. We found the numbers we used to February 2015).
terminate the reduction of K based on samplings from 7. YZang“A review on image segmentation techniques
2017

game play images and manipulations on the results, with remote sensing perspective” IAPRS Vienna,
general pictures are not supposed to work this way. Austria, July 5–7, 2010, IAPRS, Vol. XXXVIII.
Year

However we wanted to emphasize the fact that no 8. Navneet Kaur et.al “ Bacteria foraging based image
representation is better than the other. In the restricted segmentation “An International Journal of
40 world of Game play images, HSV based segmentation Engineering Sciences ISSN: 2229-6913 Issue July
did much better than the RGB version. The fact that we 2012, Vol. 6.
Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I

can ignore the pixels intensity level opened up a way to 9. Rajeshwar Dass et.al “Image Segmentation
clearly segment the picture based on hue only, which, in Techniques” IJECT Vol. 3, Issue 1, Jan. - March
this scenario specifically is very good. HSV 2012 [7] Y. Zhong, K. Karu et.al, "Locating Text in
representation is not better generally all the time, Complex Color Images," Pattern Recognition, vol.
because for real world images, the RGB representation 28, no. 10, pp. 1523-1535, October 1995.
gave us better outcomes which show us about the 10. Pena M, Barbakh W, Fyfe C (2007).. "Principal
importance of having different representations of the Manifolds for Data Visualization and Dimension
same object. Those methods we implemented that did Reduction", chapter TopologyPreserving Mappings
not get the best performance. We tried to get the right K for Data Visualization, pages 132–152. Springer,
but it was so much difficult for us to detect it. We only 2007. ISBN 978-3-540-73749-0.
worked with RGB color pace and HSV color space. We 11. McAndrew A (2004). "An Introduction to Digital
tried to get best result with RGB and HSV color space Image Processing with Matlab Notes for SCM2511
with our automated K-Means algorithm but all time we Image Processing 1", School of Computer Science
did not get best result from our proposed algorithm. and Mathematics, Victoria University“
Not only that, it was so much difficult to get the right K 12. http://www.numericaltours.com/matlab/multidim_1_
from our proposed algorithm. In future we will try to color/”, Access Date: 05-04-2017.
modify our proposed automated K-Means algorithm to 13. Marques, Blanca NP, Pina P (2005). "Pattern
get the accurate K all time at all color space. Now we Recognition and Image Analysis" Singh, S.; Singh,
only worked with two color space but in future we will M.; Apte, C.; Perner, P. Estoril, Portugal, (2005),
like to work with more color space like CMYK, L*a*b, ISBN: 978-3-540-28833-6.
YCBCR, HSL etc. after work with these color spaces we 14. Kardi Teknomo – K Mean Clustering Tutorial.
can easily compare which will be best for image 15. Lloyd SP (1957.: "Least squares quantization in
segmentation. pcm." Technical note, Bell Laboratories, 1957.
Published in 1982 in IEEE Transactions on
References Références Referencias Information Theory 28, 128-137.
16. Kardi Teknomo – K Mean Clustering Tutorial,
1. Gudrun J. Klinker, Steven A. Shafer and Takeo
Teknomo, Kardi. K-Means Clustering Tutorials.
Kanade, “A Physical Approach to Color Image
http:\\people. revoledu.com\kardi\ tutorial\kMean\
Understanding,” International Journal of Computer
17. Krishna K, Narasimha MM (1999). “Genetic K-
Vision, Vol. 4, No. 1, pp. 738, 1990.
.
means algorithm,” IEEE Trans. on Systems, Man
2. Anil Z Chitade “Color based image segmentation and Cybernetics, Vol. 29,1999, pp. 3. Linda GS,
using kmeans clustering” International Journal of George CS (2001). “Computer Vision”, pp 279-325,
Engineering science and Technology, Vol. 2(10), New Jersey, Prentice-Hall, ISBN 0-13-030796-3.
2010, 5319-5325. 18. Barbakh w, Crowe M, Fyfe C (2006).. "A family of
3. R. C. Gonzalez and R. E. Woods, Digital Image novel clustering algorithms. In 7th international
Processing. Addison-Wesley, 1993. conference on intelligent data engineering and
4. Tan, Steinbach, Kumar "Introduction to Data Mining", automated learning", IDEAL2006, pages 283–290,
4/18/2004 . University of Burgos, Spain, September 2006.
5. C. Mythili and V.Kavitha "Color Image Segmentation Springer. ISSN 03029743 ISBN-13978-3-540-
using ERKFCM", International Journal of Computer 454854.

© 20 17 Global Journa ls Inc. (US)

 Color Image Segmentation using Automated K-Means Clustering with RGB and HSV Color Spaces

19. M. Veloso, E. Winner, S. Lenser, J. Bruce, and T.

Balch. Vision-Servoed Localization and Behaviors
for an Autonomous Quadruped Legged Robot
Artificial Intelligence Planning Systems, 2000.
20. Newton Laboratories. Cognachrome image capture
device. http://www.newtonlabs.com, 1999.
21. T. A. Brown and J. Koplowitz. The weighted nearest
neighbor rule for class dependent sample sizes.
IEEE Transactions on Information Theory, pages
617–619, 1979.
22. C. E. Brodley and P. E. Utgoff. Multivariate decision

2017
trees. Machine Learning, 1995.
23. Adams, R. and L. Bischof, Seeded Region Growing.

Year
IEEE Trans. Pattern Analysis and Machine
Intelligence, 1994. 16: p. 641-647. 41
24. Cheng, H.D., et al., Color Image Segmentation:

Global Journal of Computer Science and Technology ( F ) Volume XVII Issue II Version I
Advances and Prospects. Pattern Recognition,
2001. 34: p. 2259-2281.
25. Comaniciu, D., V. Ramesh, and P. Meer. Real-Time
Tracking of Non-Rigid Objects using Mean Shift. in
IEEE Conf. Computer Vision and Pattern
Recognition. 2000: p. 142-149.
26. Y. Li and Y. Shen, “Robust image segmentation
algorithm using fuzzy clustering based on kernel-
induced distance measure,” in Proceedings of the
International Conference on Computer Science and
Software Engineering, 2008, vol. 1, pp. 1065-1068.
27. D. E. Ilea and P. F. Whelan, “Color image
segmentation using a spatial k-means clustering
algorithm,” in . Proceedings of the Irish Machine
Vision & Image Processing Conference 2006 (IMVIP
2006), Dublin City University, 2006, pp. 146–53.
28. N. R. Pal and S. K. Pal, “A review on image
segmentation techniques,” Pattern Recognition, vol.
26, no. 9, pp. 1227-1294, 1993.
29. H. C. Chen and S. J. Wang, “Visible color
difference-based quantitative evaluation of colour
segmentation,” in Proceedings of the IEEE -72-
Conference on Vision Image and Signal processing,
2006, vol. 153, no. 5, pp. 598-609.

© 2017 Global Journals Inc. (US)

Global Journals Inc. (US) Guidelines Handbook 2017

www.GlobalJournals.org