1 Page 1-13 © MAT Journals 2018. All Rights Reserved
Journal of Radio and Television Broadcast
Volume 3 Issue 3
Modified IDCP Technique for Accurate Image Defogging
1
Bulbul Bandil,
2
Vikas Mahor,
3
Vandana Vikas Thakare
1
MTech Student,
2,3
Assistant Professor
1,2,3
Department of Electronics and Communication Engineering,
MITS, Gwalior, Madhya Pradesh, India
Email:
1
bulbulbandil@gmail.com,
2
vikas.mahor@ymail.com,
3
vandanavt_19@rediffmail.com
Abstract
Digital Image is an imperative part in the explanation and analysis of data, which is in the
digital type. Images and videos of outside scenes are generally affected by the bad weather
environment such as haze, fog, mist etc. So image has bad visibility of the scene caused by the
lack of quality. In current scenario the defogging techniques are not used practically in real
time systems such as cases of train accidents due to fog. This paper exhibits a study about
various image defogging techniques to eject the haze from the fog images caught in true
world to recuperate a fast and enhanced nature of fog free images. The paper also presents a
novel modified IDCP technique for efficient image defogging. In proposed work, we enhance
digital images by applying modified technique model i.e. M-IDCP (Modified-Improved Dark
Channel Prior) Technique which provide a superior quality picture with clear visibility and
distinctive color. Toward the end, we remove the first defogging image and look at them
based on their Peak Signal Noise Ratio (PSNR), Mean Square Error (MSE) parameters.
Result obtained from modified model (IDCP- DCP- HE-CHAHE) showed improved
performance in term of estimation of air-light , sky regions become bright and smoother and
halo effect is also reduced in robustness PSNR and MSE than that of (DCP- HE-CHAHE) in
addition to the quality of recovered defogged image. With this proposed work defogging
parameters as PSNR and SNR have been increased by more than 20 percentage as
compared to other literature.
Index Terms: Image Defog; DCP; IDCP; CLAHE; PSNR; MSE
INTRODUCTION
Image processing (IP) systems increment
the estimation of a picture from the defiled
picture. Outdoor scene images captured in
the bad weather are often offended because
of the entity of the haze, fog, mist, or other
media. And other bad environment by
such as fog, haze, rain and snow is the
main reason of image degradation which
have poor quality. During Fog, The light
gets spread out before reaching the camera,
when we take an image using a camera due
to a little impurity in the atmosphere .So
study on image defogging is very
necessary for IP and target recognition.
Image enhancement technique enhances
the opposite of haze image but it lead to
loss of information in image. Image
Restoration (IR) deals to the physical
method of images in fog. Picture will be set
in the debasement of haze and cloudiness.
At last, the corruption method is utilized to
make the fog-free image. [1][3]
In mostly systems, related to image
visibility degradation [7] can be hard
which run under a wide range of
environment conditions, including outdoor
object recognition systems, remote sensing
systems, smart transportation systems such
as recording of travelling vehicle data and
traffic observation system [11]. By the
presence of floating small particles in the
atmosphere, haze is produced which is
called aerosols. This is able to grasp and
spread the beams of light. Immediately in
digital vision, image dehazing has turn out
to be a major research problem. The
brought to you by CORE View metadata, citation and similar papers at core.ac.uk
provided by MAT Journals