Ijartes v1-i2-008

International Journal of Advanced Research in Technology, Engineering and Science (A Bimonthly Open Access Online
Journal) Volume1, Issue2, Sept-Oct, 2014.ISSN:2349-7173(Online)
Non Uniform Background Illumination Removal
(NUBIR) from Microscopic Images
Monika Sharma1, Ashok Kumar2
________________________________________________
ABSTRACT-Non Uniform Illumination results in
diminished structures and inhomogeneous intensities of the
image due to different texture of the object surface and
shadows cast from different light source directions. In case
of biological images, this effect is adverse. Various
techniques such as edge detection, segmentation, and
contrast enhancement using Histogram Equalization could
not differentiate between some of the particles and their
background or neighboring pixels. The paper is aimed to
remove these problems in microscopic image processing by
removing the problem of non-uniform background
illumination from the image using Adaptive Histogram
Equalization, Morphological Opening and Edge detection
techniques for particle analysis, on comparative study is
done and a new algorithm is proposed for removing the
problem of non-uniform background illumination(NUBIR)
in biological images such as visualizing and estimating the
number of particles in microscopic images and to transform
the input images to its indexed form with maximum
accuracy involving morphological openings and structuring
element design using Morphological Opening.
_______________________________________________
KEYWORDS: Morphological opening, thinning,
skeletonization, Histogram Equalization, Thresholding,
Structuring Element.
_________________________________________________
1. INTRODUCTION
Image processing is used to modify pictures to improve them
(enhancement, restoration), extract information (analysis,
recognition), and change their structure (image editing,
composition). Images can be processed by photographic,
optical, and electronic means, but image processing by the use
of digital computers is the most common method because
digital methods are fast, precise and flexible. Image
processing technology is used by planetary scientists to
enhance images of Mars, mercury, or other planets. Doctors
use this technology to manipulate CT scans and MRI images.
For human viewing, image enhancement improves the quality
of images.
________________________________________________
First Author’s Name: Monika Sharma, Electronics & Communication
Department, ACE & AR, Mithapur, India
Second Author’s Name: Ashok Kumar, Electronics & Communication
Department, ACE & AR, Mithapur, India
___________________________________________________________
Removing noise and blurring, increasing contrast, and
revealing details are examples of enhancement operations.
The term morphology means form and structure of an object.
Sometimes it refers to the arrangements and inter-relationships
between the parts of an object. Morphological
operations are related to the shapes and digital morphology is
a way to describe and analyse the shape of a digital object.
Morphology, in biology relates more directly to shape of an
organism such as bacteria. Morphological opening is a name
specific technology that creates an output image such that
value of each pixel in the output image is based on a
comparison of the corresponding pixel in the input image with
its neighbours. One can construct a morphological operation,
by choosing the size and shape of the neighbourhood that is
sensitive to specific shapes in the input image. Morphological
functions could be used to perform common image processing
tasks, such as contrast enhancement, skeletonization, noise
removal, thinning, filling and segmentation.
2.NON-UNIFORM BACKGROUND ILLUMINATION
AND EFFECTS
Non-uniform illumination can have many sources: aging
filaments, contaminated apertures, faulty reference voltages,
or non-uniform support film fabrication. Subtle electron
illumination asymmetries are more evident at moderate-to-low
magnifications and are often inadvertently enhanced by digital
contrast adjustment. The intensity inhomogeneity problem
observed in MRI is similar to this Effect. The MRI Intensity
inhomogeneity problem is manifested as a slowly varying
multiplicative effect in the acquired images. Similarly, the
non-uniform illumination can be modelled as a multiplicative
effect. The observed image is given as
: = : : + : (1)
Where s is the true signal, I is the non-uniform illumination
field and n is additive noise. The I-field does not have any
high frequency content; in other words, it varies slowly over
the image. Non-uniform illumination removal effect is
important for later processing stages such as image
registration based on correlation metrics and segmentation
based on intensity thresholding. Particle Analysis is a
technique that helps to compute the details of the components
present in the image, their shape, size (area) and number and
other characteristics of the particles or objects present in an
image. This problem is severe in case of microscopic images
captured for the purpose of bio-medical research where it is
All Rights Reserved © 2014 IJARTES Visit: www.ijartes.org Page 23

difficult to find out the exact shape, size and number of
microscopic particles due to non-uniform illumination and
sensitivity to even small fluctuations in light.
Stars Staphylococcus aureus Pencillium muold
Figure 1: Grey-Scale Image showing different microscopic images having
non- uniform texture, brighter in the certain portions and somewhere darker.
So, three different microscopic images are taken with non
uniform background. So, the technique used would be to make
an algorithm to finally examine every particle of the image, so
that every object in the image can be clearly seen, and remove
any of the problems such as less brightness, non-uniform
illumination etc. that make it difficult to differentiate between
the particles on the microscopic image in figure 1 as (a) of
stars, (b) shows staphylococcus aureus bacteria and
trichophyton rubrum and figure (c) is penicillium fungi
growing on bread. Various techniques and common
approaches to solve the problem of particle identification are
Histogram Equalization, Boundary detection, Edge Detection,
Segmentation, Linear Filtering, Morphological operations:
Dilation and Erosion etc. Due to the problem of non-uniform
illumination in the background of the image, most of these
techniques alone fail to accurately determine the objects real
boundaries because of which most of the particles appear to be
either dark or light in an image and using techniques such as
histogram equalization, edge detection, segmentation and
general image processing algorithms based on ‘region of
interest’ could not differentiate between some of the particles
and their background or neighbouring pixels and boundaries
and shapes of the resulting object changes. So, advanced
image processing tools have to be used for maximum
accuracy of the results and to identify the particles accurately
from the image without even missing a single object.
3. TECHNIQUES FOR NON-UNIFORM
BACKGROUND REMOVAL AND RESULTS
Various common techniques such as Histogram equalization
and edge detection have been studied by editing the input
picture of microscopic particles as shown in figure 1. Related
problems with these existing technologies are studied on the
presence of non-uniform illumination field in the background
of the image and an algorithm based on morphological
opening have been studied in order to remove the problems of
non-uniform background by background approximation
techniques. Available techniques are listed below.
3.1 HISTOGRAM EQUALIZATION AND CONTRAST
ENHANCEMENT
Histogram of an image represents the relative frequency of
occurrence of grey levels within an image. Histogram
modelling techniques modify an image so that its histogram
has a desired shape. Histogram equalization is used to enhance
the contrast of the image such that it spreads the intensity
values over full range. Under Contrast adjustment using
histogram equalization, overall darkness or lightness of the
image is changed, i.e. in this technique, pixel values above a
specified value are mapped to white and pixel values below
specified values are mapped to. Here we had used the results
for image of stars only. But the algorithm is implemented on
the remaining two images. The histogram plot of original
image of stars is shown in figure 2.
Fig 2: Histogram plot of the Input Image.
Fig 3: Histogram equalization (dynamic range)
After performing histogram equalization on the original
image, the HE plot are shown below in figure 3 and the
resultant image is shown in figure 4.

Fig 4: Resulting Image after Histogram equalization for contrast enhancement
on the input image
As indicated in figure 4, it is clear that histogram equalization
technique can’t be used for images suffering from non-uniform
illumination in their backgrounds. Above histograms
for the two techniques indicate that the dynamic range for the
entire image is though improved but the amplitudes for
various pixels near the center of the image with light
backgrounds have been amplified resulting in excessive
brightness near the particles present in specified locations
making the resulting image unsuitable for Particle
identifications and analysis.
Fig 5: Approximated non-uniform background image and surface
approximation background image extraction using histogram equalization
Approximated background image estimated by the technique
of histogram equalization and its surface approximation is
shown in fig.5.The surface display [0, 0] represents the origin,
or upper left corner of the image. Surface approximation
provides the highest pixel values as a curvier portion and
lower pixel values as a flat region. But the above region
depicts a lot of irregularities in the background attained due to
a lot of dark portion attained at the top of the image in the
background. It is clear from the above plotting that histogram
equalization alone could not be able to create an image of
uniform background from non-uniform background due to the
addition of large amplitude values to the lighter regions
around the objects in case of particle analysis and hence
results in faulty calculations at the end to determine each
particle individually.
3.2 MORPHOLOGY
Mathematical morphology (MM) is a theory and technique for
the analysis and processing of geometrical structures based
on set theory, lattice theory, topology, and random functions.
MM is also foundation of morphological image processing,
which consists of a set of operators that transform images
according to the above characterizations. Morphological
operations include dilation and erosion. Erosion removes the
extra pixels from the specified areas and on the other hand
dilation adds pixels to the boundaries of objects in an image.
The erosion and dilation on the image of girl is shown in
figure 6.
(a) (b) (c)
Fig. 6: (a) original image, (b) eroded image, (c) dilated image
4. PROPOSED ALGORITHM AND WORK FOR NON-UNIFORM
ILLUMINATION REMOVAL FOR
PARTICLE ANALYSIS USING MORPHOLOGY
We have proposed an algorithm with morphological opening
at first to first estimate the background of the image and then
remove the non- uniform background illumination. This is
done by creating a structuring element of the size and shape
similar to the particles present in the image (in this case, disk
shaped) and morphological opening of the image with this
structuring element. Background approximation have been
taken as the criteria to determine the close proximity to the
non-uniform background extraction using various techniques
such as Linear Filtering, Histogram Equalization and our new
technique based on morphological processes and successive
dilation and erosion followed by contrast enhancement for the
accurate particle extraction for lateral image processing.

5. RESULTS
Image commands from Image Processing toolbox is used for
computing the background of the images and enhancing the
contrast, thresholding and computing the object statistics
present in the image.
Fig 7: Non-Uniform background extraction by morphological operations in
the above image
SURFACE BACKGROUND APPROXIMATION
The methodology used for solving the problem is estimating
accurate background approximation as a surface to extract the
non-uniform background from the image and then
constructing the new image by subtracting this estimated
background from the original image. In the surface display [0,
0] represents the origin, or upper left corner of the image.
Surface approximation indicates the highest pixel values as a
curvier portion and lower pixel values as a flat region as
shown in fig. 8. In the surface display, [0, 0] represents the
origin, or upper left corner of the image. The highest part of
the curve implies that the highest pixel values of background
(and consequently input image in figure 1) occur near the
middle rows of the image. The lowest pixel values lies at the
bottom of the image and are represented in the surface plot by
the lowest part of the curve. This background approximation
obtained is exactly similar to the original non-uniform
background field and is a uniform 3-D graph with no-sudden
changes in the surface plot as it was in case of other
techniques such as histogram equalization.
Fig.8: Background approximation as a surface in 3-d view
Output image obtained after morphological opening is the
difference of this background approximation from the original
image with removal of non-uniform background problems.
Further after the final image is obtained, there is still a
problem of noise in the image that is extracted by histogram
equalization and contrast adjustment techniques. The
modification performed uses the image enhancement after the
removal of background illumination for the lateral stages.
Resulting image after subtracting the non-uniform
illumination field from the original image results in the
required image with uniform background that is suitable for
particle analysis as shown in figure 9.
(a) ( b) (c)
Fig. 9:
a) Original Image with Non-Uniform Background.
b) Non-uniform Background extraction from original image using
morphological opening and successive erosion and dilation
techniques and structuring element approach.

c) Iout = I – B, where Iout is the image obtained after the removal of
non-uniform background (B) from original image (I) uniform
background throughout the image.
We observe after removal of non uniform illumination, that
the resulting image has the problem of less brightness. To
remove these problems, we performed image enhancement
techniques at the output of the image including the contrast
and brightness adjustment and finally thresholding was done.
Images obtained after the image enhancement step results in
the final images as shown in figure 10 that is computed to be
suitable for applications in particle analysis.
Stars Staphylococcus aureus Pencillium muold
Fig.10: Final Image obtained for Particle Analysis application with non-uniform
illumination removal and contrast enhancement obtained with new
algorithm
Histogram plot of images have been compared and results
have justified in following figure x, y of figure 11
correspondingly.
Fig.11.( x) Histogram plot after the removal of non-uniform background
Fig.11.(y) Histogram plot of the final image indicating uniform distribution of
graph
6. PSEUDO COLOR IMAGES
Generally, eye cannot distinguish more than about 2 dozen
grey levels in an image. To enhance gray level variations and
make them more obvious, these are converted to pseudo
coloured, where each gray scale (generally at least 256 levels
for most displays) is mapped to pseudo colour. Pseudo colour
Image Processing is a process of assigning colours to grey
levels based on specific criterion. In other words, pseudo
colour image processing involves mapping from a single
colour (monochrome) image into three channel colour image.
The pseudo colour image of all the images is formed by the
use of pseudo command. By this we can see a grey scale
image into coloured form. Pseudo colour images of input
images are shown in figure no. 12 as (a), (b) and (c).
Stars Staphylococcus aureus Pencillium mould
Fig. no.12: Pseudo colour images of micro particles as (a), (b) and (c).

7. PARTICLE DATA ANALYSIS
In this the characteristics of particles are computed like the no.
of particles or the largest particle etc. The region props
function returns three commonly used measurements: area,
centroid (or centre of mass), and bounding box. The smallest
rectangle is represented by bounding box that can contain a
region. Table no. 1 shows the particle data of all the images.
Particle
data
Image
1(100)
Image
2(51)
No. of
objects
797 1755 1119
Area 5 1 3
Bounding
26.60000
14
box
9.80000
333
Max. Area 65 7637 44633
Largest
Empty
particle
matrix
362 Empty
Mean of
all
Particles
4.5772 53.6991 221.0956
Table No. 1: Showing Particle Data
Image
3(72)
138
177
matrix
8. COMPARISON
In order to remove the problem which was not removed by the
previous algorithm, we performed image enhancement
techniques at the output of the image including the contrast
and brightness adjustment and finally thresholding was done.
On comparing the pervious algorithm output with the new
proposed algorithm (NUBIR) output we found that the images
of stars, bacteria and fungi get improved with 16.4%, 1.3%
and 5% respectively as shown in table 2.
Microscopi
c
Images
Stars Staphyloco
ccus auerus
Pencillium
Intensity
values
Grey level
Morp
holog
y
(prev
ious
algor
ithm)
Morp
holog
y
with
thres
holdi
ng
(NU
BIR)
Mor
pho
log
y
(pre
vio
us
algo
rith
m)
Mor
phol
ogy
with
thres
hold
ing
(NU
BIR)
Mor
phol
ogy
(pre
viou
s
algo
rith
m)
Mor
phol
ogy
with
thres
hold
ing
(NU
BIR)
0
50
150
250
0
700
70
10
70
1500
50
10
400
0
170
0
500
100
0
6250
2250
1000
250
0
800
0
400
0
500
1600
0
4000
2000
500
Variations
18.2
34.6
18.
2
19.5
40
45
%
improveme
nt
16.4% 1.3% 5%
Table No.2: Comparison between Original Images and Output Images
9. CONCLUSIONS AND FUTURE WORK
It has been concluded that techniques such as edge detection,
segmentation, and contrast or brightness enhancement using
histogram equalization could not differentiate between some
of the particles and their background or neighbouring pixel.
However the proposed technique produces optimum results.
Also we find pseudo colour images of the original images. We
performed image enhancement techniques at the output of the
image including the contrast and brightness adjustment and
finally thresholding was done. Also we calculated the
characteristics of each particle, compute its area and to show
results in area based statistics and histogram equalization.
Thus we concluded on comparing morphology based

algorithm and NUBIR algorithm that the microscopic image
of stars, staphylococcus bacteria, and pencillium fungi, get
improved with 19%, 1% and 5% respectively as shown in
table 2.
In future the proposal is to compute all the details directly on
the image without computing its binary equivalent at any
stage which then will cover all the possible cases of normal
contrast enhancement using histogram equalization as well as
all the edge detection and boundary detection techniques
normally used without calculating gray thresh of the image.
REFERENCES
[1] Emerson Carlos Pedrino ,et al.“A Genetic Programming Approach to
Reconfigure a Morphological Image Processing Architecture”, may,2011
pp1-10.
[2]Yadong Wu,et al.”An Image Illumination Correction Algorithm based on
Tone Mapping”, pp245-248, 2010.
[3] M Rama Bai,“A New Approach For Border Extraction Using
Morphological Methods”,2010, pp3832-383.
[4] Kevin Loquin,et al. “Convolution Filtering And Mathematical
Morphology On An Image: A Unified View”,2010, pp1-4.
[5] Przemysław Kupidur,”Semi-automatic method for a built-up area intensity
survey using morphological granulometry”, 2010,pp271-277.
[6] Komal Vij,et al. “Enhancement of Images Using Histogram Processing
Techniques” ,2009, pp309-313.
[7] David Menotti, Laurent Najma and Jacques Faco “Contrast Enhancement
in Digital Imaging using Histogram Equalization”, 2008,pp.1-85.
[8]Aishy Amer “New binary morphological operations for effective low-cost
boundary detection”, International Journal of Pattern Recognition and
Artificial Intelligence,2002, pp.1-13.
[9] M Rama Bai “A New Approach For Border Extraction Using
Morphological Methods”, International Journal of Engineering Science and
Technology, 2010, 2(8), pp.3832-3837.

Ijartes v1-i2-008

Recommended

Recommended

More Related Content

What's hot

What's hot (19)

Similar to Ijartes v1-i2-008

Similar to Ijartes v1-i2-008 (20)

More from IJARTES

More from IJARTES (20)

Recently uploaded

Recently uploaded (20)

Ijartes v1-i2-008