ABSTRACT provides valuable information to doctors for diagnosis of

ABSTRACT

WBC detection is most important detection of various
kinds of diseases in human body as it provides valuable information to doctors
for diagnosis of diseases. The detection of WBCs and proteins through images is
fast and cheap method as there is no special need of equipment for lab testing..
Image processing tools are used on the microscopic blood images. We will be
focusing on the white blood cell in the blood images.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

1.     
INTRODUCTION

Blood
cell image has four components: plasma, red blood cells , white blood cells,
platelets. Out of these four components red blood cells and white blood cells
are mostly represented. Here we will be focusing on the white blood cells also
called as leukocytes, they are easily identified as their nucleus is darker
than the red blood cells. The leukocytes cells contain neutrophils, basophils,
eosinophils, monocytes and lymphocytes. Out of these neutrophils, basophils and
eosinophils contain granules whereas monocytes and lymphocytes do not contain
granules. Thus we can distinguish between them, not only according to shape or
size, but also thanks to the presence of granules
in the cytoplasm and also by the number of lobes in the nucleus. White blood
cells are an important part of our immune system. They help our body fight
antigens, which are bacteria, viruses, and
other toxins that make us sick.

 

 

Fig 1 microscopic blood image

 

2. DE ALGORITHM

DE (Differential Evolution) algorithm nowadays is
widely used for detection of WBCs or RBCs. Direct search algorithm mainly aims
for optimizing global multimodal functions. DE algorithm provides the exchange
of information among several solutions using mutation operator. The version of
DE algorithm used in this work is known as rand-to-best/1/bin or “DE1″1.

3. ELLIPSE DETECTION

In order to detect ellipse in images, the preprocessing
of images is done. In preprocessing edge detection algorithm is used which
leads to an edge map image. The (x,y)
coordinates of each edge pixel p are stored inside the edge vector P. An
ellipse is defined by five points as the characteristics of line are defined by
atleast two points. Consider five points that passes through an ellipse p1, p2,
p3, p4, p5. Gathering a set of five simultaneous equations which are linear in
the five unknown parameters a, b, f, g and h:

     ax^2+2hxy+by^2+2gx+2fy+1=0.

Conditions for relative
position of any point (x,y) for an ellipse is given as:

                   0
if (x,y) is outside the ellipse                             boundary

G(x,y) is a
function that verifies the pixel existence in (x,y) and is defined as:

             1 if the pixel (x,y) is an edge
point

G(x,y)=0 otherwise.

4. WBC DETECTION

Segmentation is performed in preprocessing of the
images. It is used to isolate white blood cells (WBCs) from red blood cells (RBCs)
and background. Color based segmentation is used in this process. The
parameters like brightness and gradients are used to distinguish WBCs. Depending
on these parameters threshold value is set. Threshold value generates the
labels to classify each pixel. Although a simple histogram thresholding can be
used to segment the WBCs, at this work the diffused expectation maximization (DEM)
has been used to assure better results 2. DEM is an expectation maximization
based algorithm which has been used to segment complex medical images 3.The
edge map is computed on the segmented image. The edge map is used to obtain
image representation. It preserves object structures. DE algorithm or detector
is applied on the edge map in order to detect ellipse. The edge map can also be
obtained using morphological edge detection. Morphological operations consist
of erosion, dilation, thickening and thinning, skeleton. Mostly erosion
operation is performed using matrix to detect edges in the original image. As
morphological edge detection is the traditional method of extracting borders
from binary images. Steps involved are as follows:                                   

·        
Color
based segmentation using threshold value.

·        
Applying
differential evolution algorithm.         

·        
Counting
of  white blood cells.

 

5.
CONCLUSION

In this paper
differential evolution (DE) algorithm is used for identification of white blood
cells. This algorithm is also useful for multiple ellipse detection. Color
based segmentation is used to differentiate between white blood cells, red
blood cells and background of the image. As this method is software oriented human
errors are reduced, less time consuming and no use of chemicals.

REFERENCES

1 R. Storn and K. Price, “Differential
evolution—a    simple and effi-cient
adaptive scheme for global optimization over continuous spaces,” Tech. Rep.
TR-95-012, International Computer Science Institute, Berkeley, Calif, USA,
1995.

2 G.
Boccignone, M. Ferraro, and P. Napoletano, “Diffused expectation maximisation
for image segmentation,” Electronics Letters, vol. 40, no. 18, pp. 1107–1108,
2004.

3G. Boccignone, M. Ferraro, and P. Napoletano, “Diffused
expectation maximisation for image segmentation,” Electronics Letters, vol.
40, no. 18, pp. 1107–1108, 2004.

 ABSTRACTWBC detection is most important detection of various
kinds of diseases in human body as it provides valuable information to doctors
for diagnosis of diseases. The detection of WBCs and proteins through images is
fast and cheap method as there is no special need of equipment for lab testing..
Image processing tools are used on the microscopic blood images. We will be
focusing on the white blood cell in the blood images.

1.     
INTRODUCTION

Blood
cell image has four components: plasma, red blood cells , white blood cells,
platelets. Out of these four components red blood cells and white blood cells
are mostly represented. Here we will be focusing on the white blood cells also
called as leukocytes, they are easily identified as their nucleus is darker
than the red blood cells. The leukocytes cells contain neutrophils, basophils,
eosinophils, monocytes and lymphocytes. Out of these neutrophils, basophils and
eosinophils contain granules whereas monocytes and lymphocytes do not contain
granules. Thus we can distinguish between them, not only according to shape or
size, but also thanks to the presence of granules
in the cytoplasm and also by the number of lobes in the nucleus. White blood
cells are an important part of our immune system. They help our body fight
antigens, which are bacteria, viruses, and
other toxins that make us sick.

 

 

Fig 1 microscopic blood image

 

2. DE ALGORITHM

DE (Differential Evolution) algorithm nowadays is
widely used for detection of WBCs or RBCs. Direct search algorithm mainly aims
for optimizing global multimodal functions. DE algorithm provides the exchange
of information among several solutions using mutation operator. The version of
DE algorithm used in this work is known as rand-to-best/1/bin or “DE1″1.

3. ELLIPSE DETECTION

In order to detect ellipse in images, the preprocessing
of images is done. In preprocessing edge detection algorithm is used which
leads to an edge map image. The (x,y)
coordinates of each edge pixel p are stored inside the edge vector P. An
ellipse is defined by five points as the characteristics of line are defined by
atleast two points. Consider five points that passes through an ellipse p1, p2,
p3, p4, p5. Gathering a set of five simultaneous equations which are linear in
the five unknown parameters a, b, f, g and h:

     ax^2+2hxy+by^2+2gx+2fy+1=0.

Conditions for relative
position of any point (x,y) for an ellipse is given as:

                   0
if (x,y) is outside the ellipse                             boundary

G(x,y) is a
function that verifies the pixel existence in (x,y) and is defined as:

             1 if the pixel (x,y) is an edge
point

G(x,y)=0 otherwise.

4. WBC DETECTION

Segmentation is performed in preprocessing of the
images. It is used to isolate white blood cells (WBCs) from red blood cells (RBCs)
and background. Color based segmentation is used in this process. The
parameters like brightness and gradients are used to distinguish WBCs. Depending
on these parameters threshold value is set. Threshold value generates the
labels to classify each pixel. Although a simple histogram thresholding can be
used to segment the WBCs, at this work the diffused expectation maximization (DEM)
has been used to assure better results 2. DEM is an expectation maximization
based algorithm which has been used to segment complex medical images 3.The
edge map is computed on the segmented image. The edge map is used to obtain
image representation. It preserves object structures. DE algorithm or detector
is applied on the edge map in order to detect ellipse. The edge map can also be
obtained using morphological edge detection. Morphological operations consist
of erosion, dilation, thickening and thinning, skeleton. Mostly erosion
operation is performed using matrix to detect edges in the original image. As
morphological edge detection is the traditional method of extracting borders
from binary images. Steps involved are as follows:                                   

·        
Color
based segmentation using threshold value.

·        
Applying
differential evolution algorithm.         

·        
Counting
of  white blood cells.

 

5.
CONCLUSION

In this paper
differential evolution (DE) algorithm is used for identification of white blood
cells. This algorithm is also useful for multiple ellipse detection. Color
based segmentation is used to differentiate between white blood cells, red
blood cells and background of the image. As this method is software oriented human
errors are reduced, less time consuming and no use of chemicals.

REFERENCES

1 R. Storn and K. Price, “Differential
evolution—a    simple and effi-cient
adaptive scheme for global optimization over continuous spaces,” Tech. Rep.
TR-95-012, International Computer Science Institute, Berkeley, Calif, USA,
1995.

2 G.
Boccignone, M. Ferraro, and P. Napoletano, “Diffused expectation maximisation
for image segmentation,” Electronics Letters, vol. 40, no. 18, pp. 1107–1108,
2004.

3G. Boccignone, M. Ferraro, and P. Napoletano, “Diffused
expectation maximisation for image segmentation,” Electronics Letters, vol.
40, no. 18, pp. 1107–1108, 2004.

 

x

Hi!
I'm Josephine!

Would you like to get a custom essay? How about receiving a customized one?

Check it out