Author(s):
Vikrant Singh Thakur, Kavita Thakur, Shubhrata Gupta
Email(s):
kavithakur@rediffmail.com
Address:
Ph. D. Scholar, Dept. of Electrical Engineering National Institute of Technology, Raipur, Chhattisgarh, India
School of Studies in Electronics Pt. Ravishankar Shukla University, Raipur, Chhattisgarh, India
Department of Electrical Engineering National Institute of Technology, Raipur, Chhattisgarh, India.
Published In:
Volume - 29,
Issue - 3,
Year - 2016
Cite this article:
Thakur, Thakur and Gupta (2016). Directional Fuzzy Edge Detection Based Modified Edge Regeneration System for Efficient JPEG Artifacts Reduction. Journal of Ravishankar University (Part-B: Science), 29(3), pp.49-59.
Journal of Ravishankar University - B,
29 (2&3) pp.49-59 (2016)
Directional Fuzzy Edge Detection Based
Modified Edge Regeneration System for Efficient JPEG Artifacts Reduction
Vikrant
Singh Thakura, Kavita Thakurb* and Shubhrata Guptac
aPh.
D. Scholar, Dept. of Electrical Engineering National Institute of Technology,
Raipur, Chhattisgarh, India.
bSchool
of Studies in Electronics Pt. Ravishankar Shukla University, Raipur,
Chhattisgarh, India.
cDepartment
of Electrical Engineering National Institute of Technology, Raipur,
Chhattisgarh, India.
Corresponding
author: kavithakur@rediffmail.com
[Received
17 August 2015; revised version 27 March 2016; accepted 2 April 2016]
Abstract.
Presently, the image and video compression are the most crucial and demanding
requirements for the visual data communication due to the channel bandwidth and
data storage limitations. Transform based coding using the Discrete Cosine
Transform (DCT), is a popular technique for the image and video compression.
However, at lower bit-rates, for the DCT based image compression, the
reconstructed images suffer from several visual distortions. Edge regeneration
is one of the most recent and important technique to suppress the visual
distortions generated by DCT compression at lower bit-rates. This paper
proposed an innovative modification of the edge regeneration technique by
increasing its initial edge prediction capability under the presence of
artifacts using directional fuzzy edge detection. The proposed JPEG artifacts
reduction system addresses all the three types of artifacts, which are common
in JPEG images: blocking, edges blurring, and aliasing. Furthermore, the
proposed system enhances the quality of the JPEG compressed images via two
stages. First, it removes blocking artifacts via boundary smoothing and guided
filtering. Then, in the second stage, it reduces blurring and aliasing around
the edges via proposed modified local edge regeneration based on directional
fuzzy edge detection technique. A sound comparison of the proposed algorithm
with other existing JPEG artifact removal algorithms has been also presented on
the basis of the two important parameters, peak signal to noise ratio (PSNR)
and mean square error (MSE). Performance evaluation illustrates that the
proposed system provides maximum PSNR, minimum MSE, and hence leads to
efficient JPEG artifact reduction as compared to the other state of the art
algorithms.
Keywords:
JPEG image artifacts removal, Blocking artifacts, Edge blurring, Edge
regeneration, Directional fuzzy edge detection, DCT, PSNR, MSE.
NOTE: Full version of this manuscript is available
in PDF.