Image Denoising Using Non Linear Filter: Mr. Vijay R. Tripathi

International Journal of Modern Engineering Research (IJMER)
www.ijmer.com Vol.2, Issue.6, Nov-Dec. 2012 pp-4543-4546 ISSN: 2249-6645

www.ijmer.com 4543 | Page

Mr. Vijay R. Tripathi
Assistant Professor IN College of Engg & Technology, Akola-444004.

Abstract: Noise in an image is a serious problem In this
project, the various noise conditions are studied which are:
Additive white Gaussian noise (AWGN), Bipolar fixed-
valued impulse noise, also called salt and pepper noise
(SPN), Random-valued impulse noise (RVIN), Mixed noise
(MN). Digital images are often corrupted by impulse noise
during the acquisition or transmission through
communication channels the developed filters are meant for
online and real-time applications. In this paper, the
following activities are taken up to draw the results: Study
of various impulse noise types and their effect on digital
images; Study and implementation of various efficient
nonlinear digital image filters available in the literature
and their relative performance comparison;

I. Introduction
Today digital imaging is required in many
applications e.g., object recognition, satellite imaginary,
biomedical instrumentation, digital entertainment media,
internet etc. The quality of image degrades due to
contamination of various types of noise. Noise corrupts the
image during the process of acquisition, transmission,
storage etc[1]. For a meaningful and useful processing such
as image segmentation and object recognition, and to have
very good visual display in applications like television,
photo-phone, etc., the acquired image signal must be noise
free and made deblurred. The noise suppression (filtering)
and deblurring come under a common class of image
processing tasks known as image restoration.
In common use the word noise means unwanted
signal. In electronics noise can refer to the electronic signal
corresponding to acoustic noise (in an audio system) or the
electronic signal corresponding to the (visual) noise
commonly seen as 'snow' on a degraded television or video
image. In signal processing or computing it can be
considered data without meaning; that is, data that is not
being used to transmit a signal, but is simply produced as an
unwanted by-product of other activities. In Information
Theory, however, noise is still considered to be
information. In a broader sense, film grain or even
advertisements in web pages can be considered noise.
In early days, linear filters were the primary tools
in signal and image processing However; linear filters
have poor performance in the presence of noise that is not
additive as well as in systems where system nonlinearities
or non-Gaussian statistics are encountered. Linear filters
tend to blur edges, do not remove impulsive noise
effectively, and do not perform well in the presence of
signal dependent noise. To overcome these shortcomings,
various types of nonlinear filters have been proposed in the
literature.

II. Median Based Filters
In order to effectively remove impulse noise as
described in while preserving image details, ideally the
filtering should be applied only to the corrupted pixels, and
the noise-free pixels should be kept unchanged. This can be
achieved by determining whether the current pixel is
corrupted, prior to possibly replacing it with a new value.
Decision-based filters correspond to a well-known class of
filters that appear to be particularly efficient to reduced
impulse noise. In this work, we propose an impulse
detection scheme by successfully combining the SM filter
with CWM filter.

PROGRESSIVE SWITCHING MEDIAN FILTER
A median-based filter, progressive switching
median (PSM) filter, is implemented to restore images
corrupted by saltpepper impulse noise. The algorithm is
developed by the following two main points: 1) switching
schemean impulse detection algorithm is used before
filtering, thus only a proportion of all the pixels will be
filtered and 2) progressive methodsboth the impulse
detection and the noise filtering procedures are
progressively applied through several iterations. The noise
pixels processed in the current iteration are used to help the
process of the other pixels in the subsequent iterations. A
main advantage of such a method is that some impulse
pixels located in the middle of large noise blotches can also
be properly detected and filtered. Therefore, better
restoration results are expected, especially for the cases
where the images are highly corrupted.

Fig. 1. A general framework of switching scheme-based
image filters.

PSM Filter

1. Impulse Detection
Similar to other impulse detection algorithms, this
impulse detector is implemented by prior information on
natural images, i.e., a noise-free image should be locally
smoothly varying, and is separated by edges [4]. The noise
considered for this algorithm is only saltpepper impulsive
noise which means: 1) only a proportion of all the image
pixels are corrupted while other pixels are noise-free and 2)
a noise pixel takes either a very large value as a positive
impulse or a very small value as a negative impulse. In this
chapter, we use noise ratio (0 1) R R s s to represent how
much an image is corrupted. For example, if an image is
corrupted by R = 30% impulse noise, then 15% of the
Image Denoising Using Non Linear Filter
pixels in the image are corrupted by positive impulses and
15% of the pixels by negative impulses.
Two image sequences are generated during the
impulse detection procedure. The first is a sequence of gray
scale images,
(0) (1) (2) ( )
( , ) ( , ) ( , ) ( , )
{ , , ,.... .....}
n
i j i j i j i j
x x x x , where the
initial image
(0)
( , ) i j
x is noisy image itself , (i , j) is position of
pixel in image, it can be 1 s i s M, 1s j s N where M
and N are the number of the pixel in horizontal and vertical
direction respectively, and
( )
( , )
n
i j
x is image after n
th
iteration.
The second is a binary flag image sequence,
(0) (1) (2) ( )
( , ) ( , ) ( , ) ( , )
{ , , ,.... }
n
i j i j i j i j
f f f f where the binary flag
( )
( , )
n
i j
f is
used to indicate whether the pixel at (i, j) in noisy image
detected as noisy or noise-free after n
th
iteration. If
( )
( , )
n
i j
f =0
means pixel at (i, j) has been found as noise-free after n
th

iteration and if
( )
( , )
n
i j
f =1 means pixel at (i, j) has been found
as noisy after n
th
iteration. Before the first iteration, we
assume that all the image pixels are good, i.e.
(0)
( , ) i j
f =0 for
all (i, j).
In the nth iteration (n= 1, 2, 3) for each pixel
( 1)
( , )
n
i j
x

we first find out the median value of the samples in a
W
D
W
D
(W
D
is an odd integer not smaller than 3) window
centered about it. To represent the set of the pixels within a
W
D
W
D
window centered about
( 1)
( , )
n
i j
x

is
( 1)
( , )
n
i k j l
x

+ +
where
, W k W W l W s s s s k s W, -Ws l s W and
W1, then we have median value of this window
( 1)
( , )
n
i j
m

is

( 1)
( , )
n
i j
m

= median (
( 1)
( , )
n
i k j l
x

+ +
)
The difference between
( 1)
( , )
n
i j
m

and
( 1)
( , )
n
i j
x

provides us with
a simple measurement to detect impulses

( 1) ( 1) ( 1)
( , ) ( , ) ( , ) ( )
( , )
,
1,
n n n
i j i j i j n
i j
f if x m T
f
otherwise

<

Where T is a predefined threshold value. Once a pixel (i, j)
is detected as an impulse, the value of
( )
( , )
n
i j
x is subsequently modified

( 1) ( ) ( 1)
( , ) ( , ) ( , )
( )
( , )
( 1) ( ) ( 1)
( , ) ( , ) ( , )
,
,
n n n
i j i j i j
n
i j
n n n
i j i j i j
m if f f
x
x if f f

=
=

=

(4.3)
Suppose the impulse detection procedure is stopped after
the N
D
th iteration, then two output images-
( )
( , )
D
N
i j
x and
( )
( , )
D
N
i j
f are obtained, but only
( )
( , )
D
N
i j
f is useful for our noise
filtering algorithm.

2. Noise Filtering
Like the impulse detection procedure, the noise
filtering procedure also generates a gray scale image
sequence,
(0) (1) (2) ( )
( , ) ( , ) ( , ) ( , )
{ , , ,.... .....}
n
i j i j i j i j
y y y y and a binary
flag image sequence
(0) (1) ( )
( , ) ( , ) ( , )
{ , ,.... .....}
n
i j i j i j
g g g . In the
gray scale image sequence, we still use
(0)
( , ) i j
y to denote the
pixel value at position (i, j) in the noisy image to be filtered
and use
( )
( , )
n
i j
y to represent the pixel value at position (i, j)
in the image after the nth iteration. In a binary flag
image
( )
( , )
n
i j
g , the value
( )
( , )
n
i j
g =0 means the pixel (i, j) is
good and
( )
( , )
n
i j
g = 1 means it is an impulse that should be
filtered. A difference between the impulse detection and
noise-filtering procedures is that the initial flag image
(0)
( , ) i j
g of the noise-filtering procedure is not a blank image,
but the impulse detection result
( )
( , )
D
N
i j
f , i.e.,
(0)
( , ) i j
g =
( )
( , )
D
N
i j
f .
In the nth iteration (n = 1; 2; .), for each
pixel
( 1)
( , )
n
i j
y

, we also first find its median value
( 1)
( , )
n
i j
m

of a
W
F
W
F
(W
F
is an odd integer and not smaller than 3)
window centered about it. However, unlike that in the
impulse detection procedure, the median value here is
selected from only good pixels with
( 1)
( , )
n
i j
g

= 0 in the
window.
Let M denote the number of all the pixels with
( 1)
( , )
n
i j
g

= 0 in the W
F
W
F
window. If M is odd, then
( 1)
( , )
n
i j
m

=
( 1) ( 1)
( , ) ( , )
{ 0, ( , ) }
n n
i j i j F F
median y g i j W W

= e
The value of
( )
( , )
n
i j
y is modified only when the pixel (i, j) is
an impulse and M is greater than 0:

( 1) ( 1)
( , ) ( , )
( )
( , )
( 1)
( , )
, 1; 0
,
n n
i j i j
n
i j
n
i j
m if g M
y
y else

= >

Once an impulse pixel is modified, it is considered as a
good pixel in the subsequent iterations

( 1) ( ) ( 1)
( , ) ( , ) ( , )
( )
( , )
( ) ( 1)
( , ) ( , )
,
0,
n n n
i j i j i j
n
i j
n n
i j i j
g if y y
g
if y m

=

=

The procedure stops after the N
F
th iteration when all of the
impulse pixels have been modified, i.e.,

( , )
( , )
F
N
i j
i j
g
=0
Then we obtain the image
( )
( , )
{ }
F
N
i j
y which is our restored
output image.
Table 1.WF(3 * 3) Median Filter,CWM Filter&PWM Filter
Noise PSNR
With
Noise
PSNR After Filtering
Median
Filter
CWM
Filter
PWM
Filter
10%
15 12 13.5 29
20%
12 12 13 27
30%
11 12 13 26
40%
9.5 12 13 24
The Experimental results are shown in
Table.2 for WF (5 * 5) PWM Filter
Table 2. WF (5 * 5) PWM Filter
Noise
PSNR with
Noisy
PSNR after
Filtering
10 15 21
20 12 19
30 11 19
40 9.5 19

The Experimental results are shown in
Table.3 for WF (7 * 7) PWM Filter

Table 3. WF (7 * 7) PWM Filter
Noise
PSNR with
Noisy
PSNR after
Filtering
10 15 14
20 12 7. 3
30 11 6
40 9.5 5.9
.

(a)

(b)

(c)

(d)

(e)

<f>

(g)

(h)

<i>
Figure 2. a, b, c, d are noisy images of Lena (512512)
corrupted by salt and pepper noise with noise density of
10%, 20%, 30%, 40% respectively and corresponding
restored image by PWM are in e, f, g, h for WF (3 * 3)

III. Conclusion
In this entire dissertation work, two different non-linear
filters are implemented and extensive experiments are
performed to obtain the results with various parameters to
assess the performance of each filter. The plot of PSNR for
these two filters is given below. The Table.1, 2 &3 below
shows the PSNR value obtain using Lena Image of size 512
x 512.
From the PSNR value mention in the simulation result it is
very clear that PWM Filter shows better performance in
suppressing impulsive noise compare to above mention
filter in suppressing impulse noise when noise exceeds from
10 % to 40 %.
& Secondly extensive experimental result show that if we
increase window size i.e. 5 * 5 & 7 * 7,we find that by
increasing the window size image get more & more
corrupted & filter is not able to suppress impulsive noise
effectively compare to when window size in filter was (3 *
3) in filter. Though simulation time required is less, which
is given in table below.

Table4: Average Run Time in Sec

Therefore from the above table it is very cleared
that as window size increases image get more & more
blurred & distorted though it requires less simulation time
compare to that when window size in filter was (3 * 3). So
mostly window size of WF (3 * 3) is preferred compare to
that of WF (5 * 5) & WF (7 * 7).

Reference
[1] R.C. Gonzalez and R.E. Woods Digital Image
Processing Second Edition.
[2] A. C. Bovik, T. S. Huang, and D. C. Munson, A
generalization of median filtering using Linear
combinations of order statistics, IEEE Tran Acoust.,
Speech, Signal Process., volASSP-31, no.6, pp.1342
1350, Dec. 1983.
[3] J. Astola and P. Kuosmanen, Fundamental of
Nonlinear Filtering, Boca Raton, CRC Press, 1997.
[4] T. Sun and Y. Neuvo, Detail-preserving median
based filters in image processing, Pattern Recognit.
Lett., vol. 15, pp. 341347, Apr. 1994

[5] I. Pitas and A. N. Venetsanopoulos, Nonlinear Digital
Filters: Principles and Applications, Boston, Kluwer,
1990.
[6] B. I. Justus son, Median filtering: Statistical
properties, Two-Dimensional Digital Signal
Processing II, T. S. Huang Ed., New York;
Springer Verlag, 1981.
[7] S-J. KO and Y. H. Lee, Center-weighted
median filters and their applications to image
enhancement, IEEE Trans. Circuits and Syst.,
vol. 38, pp. 984-993, Sept. 1991.
[8] J.-H. Wang, Rescanned minmax center-
weighted filters for image restoration, Proc.
Inst. Elect. Eng., vol. 146, no. 2, pp. 101107,
1999.
[9] Tri-State Median Filter for Image Image Denoising,
T Chen, K-K.Ma, and L-H Chen IEEE
TRANSACTIONS ON IMAGE PROCESSING,
VOL. 8, NO. 12, DECEMBER 1999.
[10] T. Kasparis, N. S. Tzannes, and Q., Detail-
preserving adaptive conditional median filters, J.
Electron. Imag, vol. 1, no. 14, pp. 358364, 1992.

Filter Used PSM
Window Size Time In
Sec
WF (3 * 3) 16sec
WF (5 * 5) 13sec
WF (7 * 7)
10sec

Image Denoising Using Non Linear Filter: Mr. Vijay R. Tripathi

Uploaded by

Copyright:

Available Formats

Image Denoising Using Non Linear Filter: Mr. Vijay R. Tripathi

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Image Denoising Using Non Linear Filter: Mr. Vijay R. Tripathi

Uploaded by

Copyright:

Available Formats

International Journal of Modern Engineering Research (IJMER)

www.ijmer.com Vol.2, Issue.6, Nov-Dec. 2012 pp-4543-4546 ISSN: 2249-6645

You might also like