Journal of Theoretical and Applied Information Technology
10
th
January 2014. Vol. 59 No.1
© 2005 - 2014 JATIT & LLS. All rights reserved
.
ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195
103
AN ADAPTIVE THRESHOLD INTENSITY RANGE FILTER
FOR REMOVAL OF RANDOM VALUE IMPULSE NOISE IN
DIGITAL IMAGES
1
S.SARAVANAKUMAR,
2
A.EBENEZER JEYAKUMAR,
3
K.N.VIJEYAKUMAR,
4
NELSON
KINGSLEY JOEL
1,3
Department of ECE, ANNA UNIVERSITY REGIONAL CENTRE, COIMBATORE
2
Director Academics, SRI RAMAKRISHNA ENGINEERING COLLEGE, COIMBATORE
4
PG Scholar, Department of ECE, ANNA UNIVERSITY REGIONAL CENTRE, COIMBATORE
E-mail:
1
sskaucbe@gmail.com,
2
ebeyjkumar@rediffmail.com,
3
vijey.tn@gmail.com,
4
joelnov20@gmail.com
ABSTRACT
A novel approach for denoising digital images corrupted by impulse noise is presented in this brief. The
proposed approach uses an efficient technique to identify pixels corrupted by random noise. This is done by
setting an intensity range for the center pixel of the selected window and checking whether the number of
pixels which fall within this range is above or below a specified threshold. If the condition for an
uncorrupted pixel fails in the selected window, the window size is increased and threshold is adaptively
changed. Experimental evaluation using MATLAB revealed that the proposed approach demonstrates
better Peak Signal to Noise Ratio (PSNR) improvement for higher noise densities when compared to the
best of the approaches used for comparison. Visual interpretation of output images revealed that our
approach preserved edges and fine details when compared to the existing algorithms.
Keywords: Random Valued Impulse Noise, Intensity Range, Soft-switching, Rank order, Peak Signal to
Noise Ratio.
1. INTRODUCTION AND RELATED WORK
Digital images are highly corrupted by
Impulse noise while during acquisition and
transmission. The impulse noise can be classified
under: salt and pepper noise and random valued
noise. The pixel which is identified as corrupted
and takes either maximum or minimum gray level
is classified as pixel corrupted by salt and pepper
noise. The corrupted pixel which takes any value
between 0 and 255 is classified as Random Valued
Impulse Noise (RVIN). Further processing of an
image for its enhancement needs this noise to be
removed; otherwise the performances of image
processing tasks such as segmentation, feature
extraction, object recognition, etc. are severely
degraded by noise [1]. Though there are various
algorithms for removal of RVIN they are not
efficient at high noise densities. So we concentrate
on design of efficient algorithm for RVIN removal
in images. Eng H.L and Ma [2] proposed a Noise
Adaptive Soft-switching Median (NASM) filter.
The filter uses a soft-switching noise-detection
scheme to identify each pixel’s characteristic,
followed by proper filtering operation. In the noise-
detection scheme, global (i.e., based on the entire
picture) and local (i.e., based on a small window)
pixel statistics are utilized in the first and the
remaining two decision-making levels respectively.
Chen and Hong Ren Wu [3] proposed an
Adaptive Center Weighted Median Filter
(ACWMF). Jianjun Zhang[4] proposed a two phase
median filter for removal of RVIN. The filter
removes impulse noise from degraded images in 2
phases. In the first phase adaptive Center Weighted
Median Filter (CWMF)[12] is used to identify
noisy pixels. In the noise removal phase he used an
iterative method based on median value.
Crnojevic et al proposed a median filter
which performs filtering operation on a pixel to
pixel basis. The proposed approach considers
median of absolute deviations to identify the
corrupted pixel. The basic principle of the proposed