A PDE-based method for ringing artifact removal on grayscale and color JPEG2000 images

As successor to JPEG, JPEG2000 aims to realize a very low bitrate. Although JPEG2000's rate distortion has been improved by approximately 30% against JPEG, as on most images compressed using overlapping transform-based algorithms, one can notice visible spurious oscillations, or ringing artifacts, around the edges of a JPEG2000-compressed image. We propose a partial differential equation (PDE) based image enhancement technique, in order to remove these ringing artifacts, while keeping the image's edges. Results are provided on grayscale and color images, showing a visible enhancement, confirmed by the PSNR increase.

[1]  R. Deriche,et al.  Les EDP en traitement des images et vision par ordinateur , 1995 .

[2]  Truong Q. Nguyen,et al.  Maximum likelihood parameter estimation for image ringing artifact removal , 2000, Proceedings 2000 International Conference on Image Processing (Cat. No.00CH37101).

[3]  Joachim Weickert,et al.  A semidiscrete nonlinear scale-space theory and its relation to the Perona - Malik paradox , 1996, TFCV.

[4]  J. Koenderink The structure of images , 2004, Biological Cybernetics.

[5]  Jitendra Malik,et al.  Anisotropic Diffusion , 1994, Geometry-Driven Diffusion in Computer Vision.

[6]  Jitendra Malik,et al.  Scale-Space and Edge Detection Using Anisotropic Diffusion , 1990, IEEE Trans. Pattern Anal. Mach. Intell..