TitelbildHauptseite TUM

Monogenic Wavelet Toolbox Plug-in for ImageJ

Author

Version history

Description

The Monogenic Wavelet Toolbox implements the monogenic wavelet frames constructed in [1]. The ImageJ plug-in expands an image (2D or 3D) with respect to a monogenic wavelet frame. This allows the investigation of amplitude and phase of the images at different detail levels. The plug-in contains two out-of-the-box applications:

Equalization of brightness

Original Corruption Mask Corrupted Image Equalization of Brightness
Barbara original Corruption Mask Barbara Corrupted Barbara equal brightness
Example 1: Original image, corruption mask (ramp function multiplied with 0.05 in the center), image corrupted by multiplying with the mask, and result of the equalization of brightness algorithm. The brightness is now balanced, in particular the difference of brightness between the dark center and its outside disappears. Even fine image structures are recovered.

Original Equalization of Brightness (with noise surpression)
Flowers Eq. of Brightness
Example 2: Equalization of Brightness for color images. Fine details are visibly clearer and the background colors are stronger.

Descreening
  • This method removes the screening effect from an image.
  • The method is parameter free, i.e. the size of the screening effect is estimated auomatically. The user does have to set any parameters manually.
CameramanCameraman screenedCameraman descreened
Example: Original image, screened image, and result of the descreening algorithm. The screening effect of the screened image is nicely interpolated and details are preserved.

Download and Installation

  • Download and install ImageJ Pfeil version 1.41 or higher.
  • Download the Monogenic Wavelet Toolbox.
  • Unzip MonogenicWaveletToolbox.zip and copy the *.jar files to the ImageJ/Plugins folder.
  • Launch ImageJ. The plug-in is now available in the ImageJ Plugins menu.
  • Note: For the processing of 3D images the ImageJ memory size should be set to at least 2000 MB (Menu Edit->Options->Memory & Threads).

Troubleshooting

  • OutOfMemoryException at executing "Equalization of Brightness": Increase the memory size for ImageJ in the menu Edit->Options->Memory & Threads to at least 50 times the image size. Alternatively, launch ImageJ from the command line via java -jar -Xms512m -Xmx1024m /pathToImageJ/ImageJ/ij.jar, where the number after -Xms gives the minumum memory size and the number after -Xmx the maxiumum memory size in megabyte.
  • Error at transformations of 3D image stacks: The 3D implementation only supports 3D stacks of dimension (2^n) x (2^m) x (2^k), where n, m, and k are integers, e.g. 128 x 64 x 32. The dimensions have to be adjusted manually beforehand.

References

  1. S. Held, M. Storath, P. Massopust, and B. Forster, Steerable wavelet frames based on the Riesz transform, Pfeil IEEE Transactions on Image Processing, Volume: 19 , Issue: 3, 2010, p. 653 - 667.
  2. S. Held, M. Storath, Monogenic Wavelet Frames, Pfeil Wavelets XIII (proceedings volume), Proceedings of Spie Volume 7446, September 2009.
  3. Poster: Wavelets for Multiscale Amplitude and Phase Decomposition

Conditions and Terms of Use

  • The Monogenic Wavelet Toolbox is a plugin for ImageJ, a general purpose image-processing package.
  • You are free to use this software for research purposes, but you should not redistribute it without our consent.
  • We expect you to include adequate citations and acknowledgments whenever you present or publish results that are based on it.
  • The Monogenic Wavelet Toolbox uses the FFT library JTransforms, Pfeil which is distributed under the terms of the MPL/LGPL/GPL tri-license.