GREYCstoration
GREYCstoration is an image regularization algorithm that processes an image by locally removing small variations of pixel intensities while preserving significant global image features, such as sharp edges and corners. The most direct application is denoising. By extension, it can also be used to inpaint or resize images. It is based on methods using nonlinear multi-valued diffusion PDEs (Partial Differential Equations) for image regularization. This kind of method generally outperforms basic image filtering techniques (such as convolution, median filtering, etc.) classically encountered in painting programs. Other image denoising plugins are available, but are not free, and the corresponding algorithms are kept secret.
| Tags | multimedia Graphics Editors |
|---|---|
| Licenses | CeCILL |
| Operating Systems | Windows Windows Windows Windows POSIX Linux |
| Implementation | C++ |
Recent releases
- 15 Jun 2008 10:57
Changes: A new image cleaning mode was added (for the command-line only). It is based on diffusion PDEs on image patch space. It gives quite interesting denoising results when parameters are set correctly, sometimes better than with the traditional "-denoise" method.
- 08 Jan 2008 10:54
Changes: This release deals with RGB/YCbCr color bases, and processes only particular channels of a hyperspectral image. Only the command line version has been updated.
- 01 Aug 2007 17:06
Changes: This version is far less memory consuming than the previous one, can deal with very big images, and has processing modes for 16-bit/components and 32-bit float-valued images, as well as multi-thread support. It also provides a new plugin for the GIMP.
- 09 May 2007 23:58
Changes: This release consumes far less memory than the previous version (2.4.2), so it can deal with very large images. There are now processing modes for 16-bit/components and 32-bit float-valued images, as well as multi-thread support. A brand new plug-in for the GIMP is provided.
- 06 Mar 2007 23:03
Changes: Processing modes for 16-bit/components and 32-bit float-valued images were added. The code was cleaned up and the footprint was reduced.
