INVESTIGATION THE EFFECT OF USING GRAY LEVEL AND RGB CHANNELS ON BRAIN TUMOR IMAGE

Dhinaharan Nagamalai et al. (Eds) : AIS, CSIT, IPPR, IPDCA - 2017
pp. 141– 148, 2017. © CS & IT-CSCP 2017 DOI : 10.5121/csit.2017.71012
INVESTIGATION THE EFFECT OF USING
GRAY LEVEL AND RGB CHANNELS ON
BRAIN TUMOR IMAGE
Ahmed B Salem Salamh
Department of Computer, Zawia University, Zawia City, Libya
ABSTRACT
Analysis the effect of using gray level on the Brain tumor image for improving speed of object
detection in the field of Medical Image using image processing technique. Specific areas of
interest are image binarization method, Image segmentation. Experiments will be performed by
image processing using Matlab. This paper presents a strategy for decreasing the calculation
time by using gray level and just one channel Red or Green or Blue in medical Image and
analysis its impact in order to improve detection time and the main goal is to reduce time
complexity.
KEYWORDS
Image Segmentation, Medical Image, Detection Time, Brain Tumor image, Image Channels,
Time Complexity
1. INTRODUCTION
Image processing is a very important field in the whole world these days and has become the
necessary field of enhancing and analyze images. The image processing technique provides
massive of methods to process and manipulate pictures in many fields, for example edge
detection, image compression, facial recognition and medical image. However, image processing
overcomes traditional use as a result of computers become more efficient and reliable. A
computer is used to making a picture with different types as the properties of images that
distinguish its extensions. On the other hand, using technique to alter and interpret real images.
The main objectives of image processing are enhancing image quality and detection of
information in order to clarify image objects in spite of sophisticated details that sometimes make
it difficult to process. It is useful to use image processing methods as the first step to digitize a
picture into an image file. The next step is applying some technique to enhance image quality. In
addition, Image processing is growing very rapidly, for example, processing of medical image is
one of the most important fields due to many reasons such as breast cancer and fracture bone
detection, which causes lots of trouble in many years due to lack of research In a global term,
image processing is one of computer applications which it uses many algorithms to process the
image. Moreover, each algorithm has advantages and disadvantages that affect on image in
different positions. Many applications have been developed in order to reach optimal technique
that enable researchers to work more efficient and fast in this field. One of the important medical
images is Magnetic resonance imaging (MRI). X-ray scanners utilize solid magnetic fields, radio
waves, and field inclinations to produce pictures of the organs in the body. For instance, MRI is
used to diagnose brain tumor in order to detect mass of abnormal tumor which is abnormal cell.

142 Computer Science & Information Technology (CS & IT)
MRI makes it conceivable to introduce many sorts of tissue differentiate by changing excitation
and reiteration times, which makes it an exceptionally adaptable instrument for imaging diverse
structures of intrigue. Because of the nature and appearance of cerebrum tumors, one MRI
arrangement is not adequate to completed portion the tumor including all its sub regions.
2. RESEARCH PAPER OBJECTIVES
This paper focuses on:
1- Analysis The effects of using gray level on Medical Image (MRI image) in order to get less
information and improve time complexity and investigate the benefits of using gray level
on the brain tumor medical image.
2- Experiments of techniques this paper use on different brain images, analysis, contrast and
discuss the results.
3- Produce the conclusion.
3. RELATED WORK
Magnetic resonance imaging (MRI) ) is a fantastic medicnal imaging, especially for brain
imaging. MRI inside the human body is useful to see the level of details. Specialists have real
specialized and monetary significance of dependable also, quick detection and classification of
brain malignancy, malignancy means in advanced stage. MRI is most proficient for the
examination of brain tumor recognition and characterization when contrasted with other imaging
procedures [1]. Brain tumor is a mass of tissue that grows out of control of the normal forces that
regulates growth [2]. The Brain tumor is the most common, occurring malignancy among human
beings [3]. The brain abnormality detection and segmentation of MRI images is an exceptionally
hard to specify and crucial to assess which is utilized as a part of surgical and restorative
arranging and appraisal. The trouble in brain picture examination is primarily because of the
necessity of identification procedures with high exactness with speed time [4]. Brain tumor is
diagnosed at advanced stages with the help of the MRI image [5]. Medical Image
Analysis provides a forum for the dissemination information of new research results in the field
of medical and biological image analysis, with special emphasis on efforts related to the
applications of computer vision, virtual reality and robotics to biomedical imaging problems [6].
MRI image segmentation is an important but inherently difficult problem in medical image
processing [7]. Image segmentation is used to separate objects from the background, and thus it
has proved to be a powerful tool in bio-medical imaging. [8]. Tumor segmentation from magnetic
resonance imaging (MRI) process is an important but time ingesting manual project performed by
doctors. Magnetic resonance imaging (MRI), computed tomography (CT), virtual mammography,
and other imaging strategy, supply an efficient way for detecting one-of-a-kind form of
tumors[4]. In clinical photograph processing, the problems of reminiscence utilization and low
execution velocity are compounded with ever-increasing sizes of information units. Normal high
decision Computed Tomography (HRCT) photo sets now consist of masses of 512x512 slices,
making up an (nearly) isotropic volume, that's quality treated as one volume for reasons of
consistency of effects over the 0.33 axis [9].
In the medical field, it is still needed searching in this area, which is very important in the life
because, it saves many people or enable specialized people to find out problems related to this as
fast as possible. Many researchers have created image processing techniques to use it in specific
system. This techniques has concentrated on advantage step of the process. For instance, using
techniques such as an effecient segmentation algorithm by Fast scanning that was fast with the

Computer Science & Information Technology (CS & IT) 143
speed [10] but is not for abnormal region of brain image. There are several basic ways to
segment any type of image. For example, Threshold Technique, Edge-Based Segmentation,
Region-based Segmentation. .However, all previous method work with convert RGB image to
grayscale as first step, which consist of value between 0 to 255. From this is good to convert
image to grayscale, but this paper proposes to process RGB image directly with extract of one
color from the RGB image without converting images. Neeraja Menon and Rohit
Ramakrishnan have proposed fast MRI Brain Image segmentation method based on Artificial Bee
Colony (ABC) [11]. The first step of their work, convert MRI image to grayscale.
P. S. Juhi and S. S. Kumar proposed Bounding box based automatic segmentation of brain
tumors using random walker and active contours from brain MRI. They introduced fast and
accurate segmentation method by segmenting the most dissimilar regions of a tumor image[12].
None of any algorithms used to process MRI image significantly with one of RGB colors of the
original image.
4. METHODS
This section shows examples of some common image processing techniques that have been used
in brain tumor for image display, analysis and human interpretation.
Experiment Steps:
1. Load color image from original MRI image.
2. Convert a gray image and read Red, Green and Blue image in order to manipulate each
one.
3. Image binarization for each Red, Green and Blue image to select objects and observe the
difference between the three images.
4. Histogram for each gray image, Red, Green and Blue image for clarification.
5. Object detection with circle for each Red, Green and Blue image.
The following flowchart shows in pictorial form how a method is finished from begin to the
end in sequential step order. Moreover, represent the five steps with flowchart is needed in
order to understand the main objective of the paper and it will clarify stages as it appear in
figure 1.

Figure 1. Clarification of experiment steps with different methods
4.1 First experiment.
In The first step, load the original MRI image with a brain tumor. The next step As we can see in
the figure 2 convert MRI image to gray image in order to process images and makes the
comparison.
Figure 2. Presents gray level information from the original MRI image which captures brain tumor
The next part shows original red, green, blue colors of the image. In this part we have a choice
to extract three colors from gray image or extract one channel from the original color image.
This paper uses the method to extract red, green, blue Sequentially and present them in order to
apply segmentation and binarization on three colors as it appears in the following figure.

Figure 3. Separate gray image (a) Red (b) Green (c) Blue
The next section is applying segmentation and binarization and function to neglect the connected
pixel in specified size in order to present object and detect brain tumor. Furthermore, this will
apply on three colors sequentially to analysis the effect of object detection. This is clear in the
figure 4.
Figure 4. Presents image binarization (a) Red (b) Green (c) Blue image
The histogram is useful to check the a distribution of discrete intensity levels or the range of
quantities in the specified graph. In the figure 5, it is clear to observe the distribution of one color
in each part red, green, blue and its has the same distribution and its presented in figure 5.
Figure 5. Presents histogram (a) Red (b) Green (c) Blue image.

The histogram in the next figure is to grayscale image comes from the original color image to
demonstrate the comparison of any color red or green or blue as a result of this there is no
difference between them as it showed in figure 6.
Figure 6. Presents histogram for the original gray image.
The object detection is very important and it takes part after applying segmentation using the
threshold method with value between 180 and 200 in order to complete segmentation and
binarization to extract a brain tumor from MRI image which is clear in the figure 7. In the three
parts of colors red, green, blue the object appears and can observe. Moreover, the goal of this
detection is to make notification about object detection on each color and observe the difference
as a result the brain tumor appears in the figure 7.
Figure 7. Shows object detection with circle(a) Red (b) Green (c) Blue image.
4.2. Second experiment.
It is valuable to test and check multiple brain tumor images , in this paper is tested , checked
,verified multiple brain images with different contrast, brightness, noise in order to achieve
detection and make the comparisons and get accurate results, This section repeats the previous
method and steps which starts with a grayscale image that comes from original MRI color image
and its clearly in figure 8, combines all parts.

Figure 8. a) Presents gray level information from the original MRI image which captures brain tumor,
b) Presents histogram for the original gray image. c) Separate Red, Green and Blue of gray image,
d) Presents histogram for each Red, Green and Blue image , e) Presents image binarization for each Red,
Green and Blue image, f) Presents object detection with circle for each Red, Green and Blue image
5. DISCUSSION THE RESULTS
The suggested method is utilized to reduce the process of information needed in the whole brain
tumor original image to detect and segment tumor in another image at the beginning of any
algorithm, the method proposed dealing directly with the orginal MRI image without need to
convert the original image to grayscale image which will lead to reduce the complexity of code
and as the result will lead to reduce the time complexity of the process image as well as minimize
the information needed by using one color. Moreover, less information and calculation when
check one of three color data. The result of the segmentation and binarization when applying on
red, green, blue color of the original image is object detection, which makes the comparison more
sufficient to prove the results . The proposed method gives very good results for different MRI
images.
6. CONCLUSION
The results prove that, improve the speed of image processing is possible by separating image
colors this paper use medical image, brain tumor as a case study to test object detection and the
result will be the same if it process separately this directly will lead to enhance of the whole
manipulation in the image. Moreover, processing color image to red or green or blue and working
with one color will bring this enhancement because of this extract on color for example Red color
from color image directly and working with it is better than change image to gray level,
furthermore it is like working with less information, less mathematic operation. The benefit of
this work will help online algorithm developers, online manipulation and even in image
uploading to test fast algorithms and online application. There is no difference in tumor detection
between gray image or using one color channel which is clear when compare histograms and

objects in this case it is better to use one channel from a color image without changing color
image to grayscale for those cases.
ACKNOWLEDGEMENTS
I would like to express my deepest appreciation to all those who provided me the possibility to
complete this research paper.
REFERENCES
[1] Ruchi D. Deshmukh & DYPIET Pimpri," Study of Different Brain Tumor MRI Image Segmentation
Techniques ", April, 2014, International Journal of Computer Science Engineering and Technology,
Vol 4, Issue 4,133-136.
[2] T. Kim & Ubiquitous," computing and multimedia applications",(2011) Berlin: Springer.
[3] Tai-Hoon Kim & Debnath Bhattacharyya, (2011) ," Analysis Brain Tumor Detection Using MRI
Image", MPCT, Hannam, university, Vol.151,pp 307-314.
[4] Madhurima Banerjee & Ranjita Chowdhury & Samir Kumar Bandyopadhyay, (2015),
"DETECTION OF BRAIN TUMOR FROM MRI OF BRAIN",Vol. 02, Issue, 12, pp. 1555-1559.
[5] S. Karthigaiselvi and T. Kalaiselvi,(2017),"Investigation of Image Processing Techniques in MRI
Based Medical Image Analysis Methods and Validation Parameters for Brain Tumor", Current
Medical Imaging Reviews, vol. 13.
[6] N. Ayache & J.S. Duncan,( 2016), Medical Image Analysis, journals., MICCAI Society .
[7] Pankaj Kr. Saini1& Mohinder Singh& Maharishi Ved Vyas,( 2015)," BRAIN TUMOR DETECTION
IN MEDICAL IMAGING USING MATLAB", Engineering College Jagadhri, Yamuna Nagar India.
[8] A. Aslam, E. Khan and M. Beg, "Improved Edge Detection Algorithm for Brain Tumor
Segmentation", Procedia Computer Science, vol. 58, pp. 430-437, 2015.
[9] Madhurima Banerjee & Ranjita Chowdhury & Samir Kumar Bandyopadhyay, (2015),
"DETECTION OF BRAIN TUMOR FROM MRI OF BRAIN",Vol. 02, Issue, 12, pp. 1555-1559.
[10] Jian-Jiun Ding & Cheng-Jin Kuo & Wen-Chih Hong, "AN EFFICIENT IMAGE SEGMENTATION
TECHNIQUE BY FAST SCANNING AND ADAPTIVE MERGING", Graduate Institute of
Communication Engineering, National Taiwan University.
[11] N. Menon & R. Ramakrishnan, 2015 , “Brain Tumor Segmentation in MRI images using
unsupervised Artificial Bee Colony algorithm and FCM clustering,” 2015 International Conference on
Communications and Signal Processing (ICCSP).
[12] P. S. Juhi & S. S. Kumar, “Bounding box based automatic segmentation of brain tumors using
random walker and active contours from brain MRI, ), 2014, ” International Conference on Control,
Instrumentation, Communication and Computational Technologies (ICCICCT.
AUTHOR
Ahmed B Salem Salamh
Department of Computer, Zawia University, Zawia City, Libya

INVESTIGATION THE EFFECT OF USING GRAY LEVEL AND RGB CHANNELS ON BRAIN TUMOR IMAGE

More Related Content

What's hot (20)

Similar to INVESTIGATION THE EFFECT OF USING GRAY LEVEL AND RGB CHANNELS ON BRAIN TUMOR IMAGE (20)

Recently uploaded (20)

INVESTIGATION THE EFFECT OF USING GRAY LEVEL AND RGB CHANNELS ON BRAIN TUMOR IMAGE