multiscaler

Decompose and recompose an image into a DCT pyramid.

Every layer of the pyramid contains as many DCT (low) frequencies as its number of pixels. There is no smoothing, so ringing is to be expected. This code is intended to allow multiscale denoising of images.

Building

To compile, use

$ mkdir build
$ cd build
$ cmake .. [-DCMAKE_CXX_COMPILER=/path/of/c++/compiler -DCMAKE_C_COMPILER=/path/of/c/compiler] [-DCMAKE_BUILD_TYPE=Debug]
$ make

To rebuild, e.g. when the code is modified, use

$ cd build
$ make

Using

To decompose an image input.png into 3 levels, respectively pyramid0.tiff, pyramid1.tiff and pyramid2.tiff, use

$ decompose input.png pyramid 3 .tiff

To recompose it (possibly after working on the single layers) on output.tiff use

$ recompose pyramid 3 .tiff output.tiff

To join two layers fine.tiff and coarse.tiff together, use

$ merge_coarse fine.tiff coarse.tiff output.tiff

Use the flag -c alpha to select the ratio for conservative recomposing.

Extras

The files decompose.py, recompose.py, and merge_coarse.py implement the above functions using the wavelets provided by the pywt package. These python programs use the piio.py package (included) for reading and writing images in the tiff format.

$ ./decompose.py input.png pyramid 3 .tiff
$ ./recompose.py pyramid 3 .tiff output.tiff