Vector Image Model to Object Boundary Detection in Noisy Images

Full text

Available online at www.ijarmate.com
International Journal of Advanced Research in Management, Architecture, Technology and
Engineering (IJARMATE)
Vol. 1, Issue 2, September 2015
All Rights Reserved © 2015 IJARMATE
13
Vector Image Model to Object Boundary Detection
in Noisy Images
S.Suryakala
1
, I.V.Sushmitha Dani
2
, I.Shibiya Sherlin
3
, S.Sheba Monic
4
, A.Sushma Thavakumari
5
, Christo Ananth
6
U.G. Scholars, Department of ECE, Francis Xavier Engineering College, Tirunelveli
1,2,3,4,5
Associate Professor, Department of ECE, Francis Xavier Engineering College, Tirunelveli
6
Abstract— A New model is designed for boundary
detection and applied it to object segmentation problem
in medical images. Our edge following technique
incorporates a vector image model and the edge map
information. The proposed technique was applied to
detect the object boundaries in several types of noisy
images where the ill-defined edges were encountered. The
proposed techniques performances on object
segmentation and computation time were evaluated by
comparing with the popular methods, i.e., the ACM, GVF
snake models. Several synthetic noisy images were
created and tested. The method is successfully tested in
different types of medical images including aortas in
cardiovascular MR images, and heart in CT images.
Index Terms—
Noisy Images, ACM, GVF snake models
I. INTRODUCTION
Image analysis is the extraction of meaningful information
from images; mainly from digital images by means of digital
image processing techniques. Image analysis tasks can be as
simple as reading bar-coded tags or as sophisticated
as identifying a person from their face. Image analysis
operations produce numerical or graphical information based
on characteristics of the original image. They break into
objects and then classify them. They depend on the image
statistics. Common operations are extraction and description
of scene and image features, automated measurements, and
object classification. Image analyse are mainly used in
machine vision applications. Computer image analysis
largely contains the fields of computer or machine vision,
and medical imaging, and makes heavy use of pattern
recognition, digital geometry, and signal processing.
It is
the quantitative or qualitative characterization
of two-dimensional (2D) or three-dimensional (3D) digital
images. 2D images are, for example, to be analyzed
in computer vision, and 3D images in medical imaging.
Image compression and decompression reduce the data
content necessary to describe the image. Most of the images
contain lot of redundant information, compression removes
all the redundancies. Because of the compression the size is
reduced, so efficiently stored or transported. The compressed
image is decompressed when displayed. Lossless
compression preserves the exact data in the original image,
but Lossy compression does not represent the original image
but provide excellent compression. Image compression may
be lossy or lossless. Lossless compression is preferred or
archival purposes and often for medical imaging, technical
drawings, clip art, or comics. Lossy methods are especially
suitable for natural images such as photographs in
applications where minor loss of fidelity is acceptable to
achieve a substantial reduction in bit rate.
Methods for lossless image compression are:
 Run-length encoding – used as default method
in PCX and as one of possible in BMP, TGA, TIFF
 DPCM and Predictive Coding
 Entropy encoding
 Adaptive dictionary algorithms such as LZW – used
in GIF and TIFF
 Deflation – used in PNG, MNG, and TIFF
 Chain codes
Methods for lossy image compression:
 Reducing the color space to the most common
colors in the image. The selected colors are
specified in the color palette in the header of the
compressed image. Each pixel just references the
index of a color in the color palette. This method
can be combined with dithering to
avoid posterization.
 Chroma sub sampling. This takes advantage of the
fact that the human eye perceives spatial changes
of brightness more sharply than those of color, by
averaging or dropping some of the chrominance
information in the image.
 Transform coding. This is the most commonly
used method. A Fourier-related transform such
as DCT or the wavelet transform are applied,
followed by quantization and entropy coding.
 Fractal compression.
In [1], Bing Li and Scott T. Acton proposed a novel
static external force for active contours, called the vector field
convolution. The VFC field is calculated by convolving a
vector field kernel with the edge map generated from the
image. Different external forces have been proposed to
improve the performance of snakes. The external forces can
be generally classified as dynamic forces and static forces.
The dynamic forces are those that depend on the snake and, as

Available online at www.ijarmate.com
International Journal of Advanced Research in Management, Architecture, Technology and
Engineering (IJARMATE)
Vol. 1, Issue 2, September 2015
All Rights Reserved © 2015 IJARMATE
14
a result, change as the snake deforms. The static forces are
those that are calculated from the image, and remain
unchanged as the snake de-forms. The main disadvantage of
this method id only high capture range images can be used
here.
In [2], Chenyang Xu and Jerry L. Prince proposed a
method called gradient vector flow (GVF). The gradient
vector flow fields are dense vector fields derived from images
by minimizing certain energy functional in a variation
framework. The minimization is achieved by solving a pair of
decoupled linear partial differential equations that diffuse the
gradient vectors of a gray-level or binary edge map computed
from the image. The GVF snake also has a large capture
range, which means that, barring interference from other
objects, it can be initialized far away from the boundary. The
disadvantage of the model is resultant value is only closer not
accurate.
In [3], Chenyang Xu and Jerry L. Prince presents a
new class of external forces for active contour model gradient
vector flow (GVF) fields, are dense vector fields derived from
images by minimizing energy functional in a variation
framework. The minimization is achieved by solving a pair of
decoupled linear partial differential equations which diffuses
the gradient vectors of a gray-level or binary edge map
computed from the image. Particular advantages of the GVF
snake over a traditional snake are its insensitivity to
initialization and ability to move into concave boundary
regions. A new external force model for snakes called
gradient vector flow (GVF). The disadvantage of the method
is performance degradation.
In [4], François Destrempes, Gilles Soulez and Guy
Cloutier proposed a new semi-automatic segmentation
method. The intima-media thickness (IMT) is a double-line
pattern visualized by echography on both walls of the
common carotid arteries in a longitudinal image. It is formed
by two parallel lines, which consist of the leading edges of
two anatomical boundaries: the lumen-intima and
media-adventitia interfaces. Semi-automatic segmentation of
the IMT on the far wall of carotid arteries in B-mode
ultrasound images is a useful tool for clinical applications.
Rayleigh distributions have been used to model the local
brightness of the speckle pattern in a B-mode image. The main
disadvantage of this method is only high contrast image can
be used here.
In [5], FrCdCric Leymarie and Martin D. Levine
proposed a explored a new segmentation technique Snakes
can be represented as energy-minimizing splines guided by
external constraint forces and image forces such as lines,
edges, subjective contours, and region homogeneities found
in the image. Furthermore, internal spline forces impose
smoothness constraints on the modeled contours. By
combining and integrating various types of information found
in an image, snakes can lead to results that are at least
comparable with other image-segmentation techniques. The
steady-state criterion for terminating the optimization
procedure is an overly strong one, leading to unnecessary
oscillations and, possibly, the snake receding from a valid
solution. The drawback of the model is, the model contains
multiple processes.
II. PROPOSED
SYSTEM
Our proposed system is edge following technique for
boundary detection in noisy images. Utilization of the
proposed technique is exhibited via its application to various
types of medical images. Our proposed technique can detect
the boundaries of objects in noisy images using the
information from the intensity gradient via the vector image
model and the texture gradient via the edge map. The
performance and robustness of the technique have been tested
to segment objects in synthetic noisy images and medical
images including prostates in ultrasound images, left
ventricles in cardiac magnetic resonance images, aortas in
cardiovascular MR images, and knee joints in computerized
tomography images.
It gives more information for searching the boundary of
objects and increases the probability of searching the correct
boundary. The magnitude and direction of the average edge
vector field give information of the boundary which flows
around an object. In addition, the edge map gives information
of edge which may be a part of object boundary. Hence, both
average edge vector field and edge map are exploited in the
decision of the edge following technique.
Fig.1. Overall System Design
The overall system design of the Edge following
technique using the average edge vector field and edge map
method is shown in the figure 5.1 with the deterministic
annealing and nonexpansive mapping. It gives more
information for searching the boundary of objects and
increases the probability of searching the correct boundary.
The results of detecting the object boundaries in
noisy images show that the proposed technique is much better
than the five contour models. The results of the running time
on several sizes of images also show that our method is more
efficient than the five contour models.

Available online at www.ijarmate.com
International Journal of Advanced Research in Management, Architecture, Technology and
Engineering (IJARMATE)
Vol. 1, Issue 2, September 2015
All Rights Reserved © 2015 IJARMATE
15
Fig.2. Input image
Fig.3. Preprocessed image
Fig.4. Edge map of the image
III. CONCLUSION
A New model is designed for boundary detection and
applied it to object segmentation problem in medical images.
Our edge following technique incorporates a vector image
model and the edge map information. The proposed technique
was applied to detect the object boundaries in several types of
noisy images where the ill-defined edges were encountered.
The proposed techniques performances on object
segmentation and computation time were evaluated by
comparing with the popular methods, i.e., the ACM, GVF
snake models. Several synthetic noisy images were created
and tested. The method is successfully tested in different
types of medical images including aortas in cardiovascular
MR images, and heart in CT images.
REFERENCES
1. Bing Li and Scott Acton T. (2013), ‘Active contour external force
using vector field convolution for image segmentation’, IEEE
Transactions on Image Processing, Vol. 16, No. 8, pp.
2096–2106.
2. Chenyang Xu and Jerry Prince L. (2011), ‘Snakes, shapes, and
gradient vector flow’, IEEE Transactions on Image Processing,
Vol. 7, No. 3, pp. 359–369.
3. Chenyang Xu and Jerry Prince L. (2010), ‘Gradient vector flow: A
new external force for snake’, in IEEE Proceedings on
Conference Computation Visual Pattern Recognition, pp. 66–71.
4. François Destrempes, Gilles Soulez and Guy Cloutier (2009),
‘Segmentation in ultrasonic B-mode images of healthy carotid
arteries using mixtures of Nakagami distributions and stochastic
optimization’, IEEE Transactions on Image Processing, Vol. 28,
No. 2, pp. 215–229.
5. FrCdCric Leymarie and Martin Levine D. (2008), ‘Tracking
deformable objects in the plane using an active contour model’,
IEEE Transactions on Pattern Analysis and
Machine Intelligence, Vol. 15, No. 6, pp. 617–634.