Sub1528

International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2013): 4.438
Volume 4 Issue 1, January 2015
www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Reconfigurable Binary Image Processor for Brain
Tumor Detection
Sibin .P .A1
, D. Chidhambaram2
1
M.Tech Student, Electronics and Communication Engineering, Hindusthan Institute of Technology, Coimbatore India
2
Assistant Professor, Electronics and Communication Engineering, Hindusthan Institute of Technology, Coimbatore India
Abstract: Nowadays medical field is looking forward for the sophisticated hardware and software VLSI implementations. The next
generation medical diagnosis requires fast and robust dedicated hardware for precise applications, MRI images are commonly used for
tumor detection and complex hardware are necessary for its better computations .This rises the necessity of dedicated hardware for
improved performance. Binary images are optimal for hardware implementation. We propose a novel method in which MRI images are
converted to binary images which allows the real time hardware implementation. VLSI implementation using binary images effectively
improves performance of the system in terms of reduction in computational complexity, hardware requirements, reduces memory space
thus efficiently reduces cost.
Keywords: binary images, morphological filtering, k-mean filtering
1. Introduction
Images are one of the basic communication strategy through
which we can convey messages in a two dimensional
platform. Digital image processing helps in extracting set of
characters that convey some messages which are helpful in
medical diagnosis. Medical images which are the outcomes
of MRI or ultra scans especially of internal organs don’t give
a good quality image which makes visual diagnosis hardly
impossible. so further complex processing hardware for
medical diagnosis.
1.1 Binary Images
Binary images which possess only two level pixels are
simpler and powerful image for digital computations. Binary
images only have foreground and background which is
represented by 1’s and 0’s only which reduces memory
requirements which allows hardware implementations. The
lack of easier conversion methodologies to convert medical
images to binary images this problem in binary image
process implementations can be easily solved by
morphological filtering .morphological operations can be
easily applied over binary images. The structuring element
of a binary image is usually small so the segmentation
operations using k-mean filtering operations for the
detection of brain tumor can be easily processed.
The morphological operations improves noise margin and
Since the pixels can only have two values, the
morphological operations are erosion (any unset pixels
within the structuring element cause the pixel to be unset)
and dilation (any set pixels within the structuring element
cause the pixel to be set). Important operations are
morphological opening and morphological closing which
consist of erosion followed by dilation and dilation followed
by erosion, respectively, using the same structuring element.
Opening tends to enlarge small holes, remove small objects,
and separate objects. Closing retains small objects, removes
holes, and joins objects.
1.2 Challenges
Medical image processing requires high precision images for
the accurate decisions but the medical imaging of the human
interior is much complex and thus even using high end
devices we don’t have a high resolution images which limits
the medical diagnosis. The quantitative and qualitative bio-
medical image processing is limited by the computational
problem related to low resolution images. Due to the
limitation of processing a low resolution images. The lack of
standard test cases and standardization further set limitations
to the image processing. The imaging techniques in the
molecular level as well as sub cellular level complicate the
processing capabilities of the system.
The major drawback for the current system is completing in
recommending images in the database for the processing.
Thus linking or external hard disk memory usage
complicates the processing capabilities and requirements.
Whatever be the processing idealisms or structures used the
limitations in visually identify and processing capabilities of
a physician limits the image processing tools also have to
consider how well the image can illustrate the current
circumstances. In many cases medical images are the proof
for a disease or a severe circumstance of a person. But the
degraded image can’t help the doctors for the decision
making.
The brain tumors are created by an abnormal cell growth in
brain the Axial MRI view is used to detect the cell growth.
Commonly a patient is diagnosed by symptoms. And its
conformed by biopsy the imaging techniques MRI/CT scans
are used to detect and localize the tumor and also to
materialize the tumor growth. MRI-images are commonly
recommended since it doesn’t involve harmful radiations.
The MRI image of patient is the only measure to localize the
tumor growth. Since the MRI images are not having a high
resolution so visual detection is not much accurate. For that
we use image processing tools to detect and localize tumor
growth. The medical MRI images s itself is not suitable for
image processing so the original image is segmented using
Paper ID: SUB1528 356

www.ijsr.net
K-mean filtering and morphological operations. These
operations preprocessing steps change the original image to
a computer suitable form here it’s also converted to binary
image so that the hardware implementation also of the
processor is possible as a simple decided hardware. In the
field of brain tumor detection different approaches have
been carried out. Sindhusree [14] have detected tumor in 3D
view. Mc jobilchrist [6] proposed a method using K-mean
clustering. K-means clustering with marker controlled
segmentation and with fuzzy c-means clustering for medical
image segmentation. The proposed methodology under goes
a two stage process, K-means clustering for the primary
segmentation secondly marker controlled segmentation to
get the finest image. To detect tumor from MRI using fuzzy
clustering technique P Vasuda [12] [proposed a technique
but it requires a more computation technique. Bin Zhang [1]
proposed binary images for morphological operations which
reduces computation and complexity and reduces hard ware
cost. Supervised segmentation method needs considerable
amount of testing data which complicate the hardware
requirements and thus increase cost Ananda resmi [13]
2. Proposed Methodology
Binary images replacing the MRI images not only allow a
fast reliable and low cost dedicated hardware. It also
improves the operational computation features since
morphological operations such as dilation, erosion are easily
processed in binary images .The complex segmentation
operations are also effectively implemented in Binary
images .here morphological filletring to K-mean clustering
algorithm for segmentation to detect the brain tumors.
Figure 1:- proposed block diagram
The grey scale MRI image is converted to binary image,
then its pasted to morphological filtering operations as a pre
processing of image then the image is clustered by using K-
mean clustering algorithm are recommended to perform a
morphological filtering to avoid clustered regions.
2.1 Morphological Filtering
A set of image processing operations that a basically done to
emphasis an image for digital image processing application
are known as morphological filtering operations. The basic
operation of morphological filtering dilation and erosion is a
comparison technique which optimizes the image for other
digital image operation such as clustering and segmentation.
In this paper the pre processing operations are carried out to
convert a MRI image to binary image thus morphological
operations can be implemented effectively.Since the dilation
and erosion operations are carried out in binary Images the
operation is assumed to setting a pixel value to one. By
comparing with the nearby pixels
2.2 K-Means Clustering
Clustering is a method of grouping of related things together
to perform similar groups. K-means is an algorithm
commonly used to cluster by defining cluster center and
finds the position of other cluster points as a relative
positioning from the cluster data. To minimize the distance
from data points to cluster center.
Algorithm
K-means [Llyods algorithm]
1. Decides the number of clusters ‘k’.
2. Initialize the center of the clusters.
3. Attribute the closest cluster to each data point.
4. Set the position of each clusters to the mean of all data
points belonging to that clusters.
5. Repeat the steps 3&4 until it converges.
3. Simulation Result
Tumor detection from an MRI image using our proposed
method. The brain tumor is located from equivalent binary
image using K-mean cluster algorithm and morphological
filtering.
Figure 2: Original MRI image with tumor
Figure 3: dilation and erosion outputs of binary image

www.ijsr.net
Using mat lab we success fully located the tumor growth in
an input MRI image using K-men clustering and
morphological filtering
Figure 4: Tumor detected binary image
4. Conclusion
Simplified computation method on a MRI image for tumor
detection is detailed above using K-means and
morphological filtering on a binary image which is extracted
from the input image. Since binary images are used for
image processing it paves the hardware implementation.
This technique leads to the reduction in existing hardware
complexity end thus reduces cost .future scope of this work
should be focused on on-chip implementation of the
hardware.
References
[1] “Reconfigurable Processor for Binary Image
Processing” Bin Zhang, Kuizhi Mei, Member, IEEE,
and Nanning Zheng, Fellow, IEEE
[2] Akansha Singh , Krishna Kant Singh, “A Study Of
Image Segmentation Algorithms For Different Types Of
Images”,International Journal of Computer Science
Issues, vol. 7,Issue 5, pp 414-417,2010.
[3] “Intelligent fpga based system for shape recognition”
Emerson C. Pedrino, Orides Morandin Jr.,Edilson R. R.
Kato, Valentin O. Roda
[4] M.C. Jobin Christ, R.M.S.Parvathi, “Segmentation of
Medical Image using Clustering and
WatershedAlgorithms”, American Journal of Applied
Sciences,vol. 8, pp 1349-1352, 2011
[5] E. C. Pedrino, O. Morandin, Jr., and V. O. Roda,
“Intelligent FPGA based system for shape recognition,”
in Proc 7th Southern Conf.Programmable Logic, 2011,
pp. 197–202.
[6] E. C. Pedrino, J. H. Saito, and V. O. Roda,
“Architecture for binary mathematical morphology
reconfigurable by genetic programming,” in Proc. 6th
Southern Programmable Logic Conf., 2010, pp. 93–98.
[7] H. Yang and A. C. Kot, “Binary image authentication
with tampering localization by embedding
cryptographic signature and block identifier,” IEEE
Signal Process. Lett., vol. 13, no. 12, pp. 741–744, Dec.
2006.
[8] H. Yang and A. C. Kot, “Pattern-based data hiding for
binary image authentication by connectivity-
preserving,” IEEE Trans. Multimedia, vol. 9, no. 3, pp.
475–486, Apr. 2007.
[9] M. Laiho, J. Poikonen, and A. Paasio, “Space-
dependent binary image processing within a 64×64
mixed-mode array processor,” in Proc. Eur. Conf.
Circuit Theory Design, 2009, pp. 189–192.
[10]J. Velten and A. Kummert, “Implementation of a high-
performance hardware architecture for binary
morphological image processing operations,”in Proc.
47th IEEE Int. Midwest Symp. Circuits Syst., Jul. 2004,
pp. 241–244.
[11]A. Zarandy, A. Stoffels, T. Roska et al.,
“Implementation of binary and gray-scale mathematical
morphology on the CNN universal machine,” IEEE
Trans. Circuits Syst. I, vol. 45, no. 2, pp. 163–168, Feb.
1998.
[12]P.Vasuda, S.Satheesh, “Improved Fuzzy C-Means
Algorithm for MR Brain Image Segmentation”,
International Journal on Computer Science and
Engineering (IJCSE), vol. 02, no.05, pp 1713-
1715,2010.
[13]Ananda Resmi S, Tessamma Thomas, ”Automatic
Segmentation Framework for Primary Tumors from
Brain MRIs Using Morphological Filtering
Techniques”, in 5th Int Conf on Biomedical
Engineering and Informatics,2012,IEEE
[14]S. Chien and L. Chen, “Reconfigurable morphological
image processing accelerator for video object
segmentation,” J. Signal Process. Syst., vol.62, no. 1,
pp. 77–96, 2011
[15]H. Hedberg, P. Dokladal, and V. Owall, “Binary
morphology with spatially variant structuring elements
algorithm and architecture,” IEEE Trans. Image
Process., vol. 18, no. 3, pp. 562–572, Mar. 2009.
[16]J.Vijay, J.Subhashini, “An Efficient Brain Tumor
Detection Methodology Using K-Means Clustering
Algorithm”, in Int Conf on Communication and Signal
Processing, 2013, IEEE.
[17]J.Vijay, J.Subhashini, “An Efficient Brain Tumor
Detection Methodology Using K-Means Clustering
Algorithm”, in Int Conf on Communication and Signal
Processing, 2013, IEEE.

Sub1528

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to Sub1528 (20)

More from International Journal of Science and Research (IJSR) (20)

Sub1528