Asian Journal of Information Technology

Year: 2007
Volume: 6
Issue: 2
Page No. 136 - 144

Reduction of Artifacts in Jpeg Images with Genetic Algorithm and Boundary Pixel Replacement

Authors : K. Sivakami Sundari and V. Sadasivam

Abstract: Many multimedia applications require image compression with high compression ratio to overcome the difficulties in dealing huge volume of image data. At high compression ratios, the error introduced by quantization of the transform coefficients produces visually undesirable patterns known as compression artifacts that dramatically lower the perceived quality of a particular image. Blocking artifacts of JPEG images and ringing artifacts of JPEG 2000 images plays crucial role in many applications. A great deal of effort has been invested in attempts to solve this problem while preserving the information content of the image. Proposed research primarily concentrates on the blocking artifacts of JPEG images and to a degree over the ringing artifacts of JPEG 2000 images. There exist three different approaches to reduce the artifacts as Preprocessing, Post processing and Transform domain techniques. Recently, attention is diverted to optimize the solution. To enhance the performance of the algorithm principally, the artifacts are to be detected. This in turn needs some metrics to measure these distortions. The metrics used commonly for measuring these distortions are Mean Square Error (MSE) and SNR (Signal to Noise Ratio). Current research computes the measure of blocking artifacts with the new parameter named as Total Blocking Error (TBE). Minimization of TBE is an indication about the elimination of the artifacts. This can be implemented in Transform domain with a modified quantisation table and filter. Efficient suppression of artifacts can be controlled by the scaling parameter in the quantisation process and by the kernel in the filtering process. Hence the problem can be stated as finding an optimal solution for the suppression of Artifacts with these two processes. Genetic Algorithm (GA) is one of the emerging optimization techniques. So far GA has not been used for the optimization of the reduction of artifacts. Hence an attempt is made to optimize the kernel of the filter and the scaling parameter of the quantizationwith GA. A spatial domain algorithm can enhance further the quality of the image by preserving fine details. A spatial domain algorithm can enhance further the quality of the image by preserving fine details. Dynamic range processing divide the image into luminance and chrominance component and converted to a reduced range with logarithmic mapping. Attenuating the magnitudes of large gradients processes gradient field of the luminance image. Solving a poisson equation on the modified gradient field preserves fine details. Finally the integrated in formations are remapped to the original dynamic range with inverse logarithm.

How to cite this article:

K. Sivakami Sundari and V. Sadasivam , 2007. Reduction of Artifacts in Jpeg Images with Genetic Algorithm and Boundary Pixel Replacement. Asian Journal of Information Technology, 6: 136-144.

Design and power by Medwell Web Development Team. © Medwell Publishing 2020 All Rights Reserved