In these pages we provide access to the material (software, publications, experimental data, presentations, test-images, etc.) related to our recent research in the field of shape-adaptive transforms for image filtering.
Examples of the adaptive-shape supports used for SA-DCT domain filtering.
The two-dimensional separable DCT, computed on a square or rectangular support, is a well established and very efficient transform in order to achieve a sparse representation of image blocks. For natural images, its decorrelating performance is close to that of the optimum Karhunen-Loeve transform. Thus, the DCT has been successfully used as the key element in many compression and denoising applications. However, in presence of edges such near-optimality fails. Because of the lack of sparsity, edges cannot be coded or restored effectively, and ringing artifacts arising from the Gibbs phenomenon become visible.
We provide powerful image filtering algorithms based on shape-adaptive transforms.
By using arbitrarily-shaped supports which are adaptive with respect to the image, the reconstructed edges are clean, and no unpleasant ringing artifacts are introduced by the fitted transform.
In particular, we exploit the low-complexity shape-adaptive DCT (SA-DCT) to realize very efficient solutions to a wide class of image restoration problems (e.g. denoising, deblurring, deringing and deblocking).
Experiments demonstrate a state-of-the-art quality of the final estimate, both in terms of objective criteria and visual appearance.
We follow the principle of “reproducible research” and make our software routines available free-of-charge for educational and non-profit scientific research, enabling others researchers to understand and reproduce our work.
Adobe Photoshop Plug-ins
The Pointwise Shape-Adaptive DCT Plug-ins for Adobe Photoshop provide filters for: - grayscale and color image denoising - JPEG deblocking and deringing
Any unauthorized use of the provided software for industrial or profit-oriented activities is expressively prohibited. By downloading any of the files contained in this site, you implicitly agree to all the terms of the TUT limited license. Please read the TUT limited license before you proceed with downloading any of the files.