MosaicSuite for ImageJ and Fiji (New)

MosaicSuite for ImageJ and Fiji

What is MosaicSuite?
Several of the image-processing algorithms developed at the MOSAIC Group for fluorescence microscopy are available as plugins for the popular free image processing software ImageJ2 or Fiji.
The first plugin which is now part of MosaicSuite was a popular 2D/3D single-particle tracking tool which can be used to track bright spots in 2D/3D movies over time. As more plugins have been added, we decided to provide them in a single, coherent package, which will also group them under a common menu point "Plugins->Mosaic" in ImageJ2 and Fiji.



How to install MosaicSuite?
MosaicSuite requires ImageJ2 or Fiji application and Java 8 to work.

If you do not have Java 8 installed on your computer (JRE or JDK) please download it and install from official Java site.

If you do not have Fiji or ImageJ installed on your computer please follow these steps:

  1. Download Fiji from Fiji site and install it. (alternatively you can download ImageJ2 here)
  2. Run Help->Update Fiji (or in case ImageJ2 run Help->Update...)
  3. Click on "Manage update sites"
  4. Find there and mark "MOSAIC ToolSuite"
  5. Apply changes and Fiji should automatically download latest release of MosaicSuite.
  6. Restart Fiji (as required) and after restart all functionality by MosaicSuite can be found in Plugins->Mosaic menu.
If you have already installed Fiji or ImageJ then go directly from point 2. If something is still not working you can try to install MosaicSuite manually as described in next section (for Java 8 dowload this JAR file).

Watch this video tutorial on how to install the software and get started using it.



Can MosaicSuite work with old ImageJ1 or Fiji with Java 6?
Yes, but it requires manual installation. First download correct jar file:

and then, copy it into the "plugins" folder of your ImageJ/Fiji installation. The "Mosaic" menu item will appear next time you restart ImageJ/Fiji.
NOTICE: There is ongoing work on transition to Java 8 for both - ImageJ2 and Fiji. More information can be found here. Therefore we strongly advice to install latest Fiji and Java 8. All new functionality added to MosaicSuite will also require Java 8 to work.



Where is MosaicSuite source code?
MosaicSuite code can be found on public MOSAIC git server.

Its code can be downloaded by following git command:
git clone https://git.mpi-cbg.de/mosaic/MosaicSuite.git

If you'd like to contribute bug fixes or new functions to any of the plugins, or are interested in using the source code in your own projects, please make sure to first download the latest version. The code is constantly evolving. If you think your additions could be useful also for other users, please send them to us and we will include them in future releases. Your contributions are highly appreciated!



Other useful information

  • all tutorials and documentation can be accessed online from: MosaicSuite tutorials.
  • all user manuals and example data can be downloaded as a packaged ZIP file from here.



Latest news:


NEW: Automatic optimal filament segmentation

The plugin can be used for a globally optimal filament segmentation of 2D images with previously unknown number of filaments. You can find plugin for segmentation in the menu "Segmentation > Filament". Presented solution can produce sub-pixel accuracy results and handle different types of image data from different microscopy modalities.

The algorithm implemented in this plug-in is described in:

X. Xiao, V. F. Geyer, H. Bowne-Anderson, J. Howard, and I. F. Sbalzarini. Automatic optimal filament segmentation with sub-pixel accuracy using generalized linear models and B-spline level-sets. Med. Image Anal., 32:157–172, 2016. (PDF)

In order to ensure financial support for our project and allow further development of this software, please cite above publications in all your documents and manuscripts that made use of this software. Thanks a lot!


Fast implicit curvature filters

Curvature filters provide geometric means of image filtering, denoising, and restoration. This amounts to solving a variational model, but the filters here implicitly do this, and are much faster. You find the filters in the menu "Enhancement - Curvature Filters". Currently, we implement Gauss curvature, Mean curvature, and Total Variation (TV) filters. The only parameters is the number of iterations, i.e., how many passes of the filter should be applied to the image. Else the filters are parameter free. A C++ implementation of these filters is also available here.

The algorithm implemented in this plug-in is described in:

Y. Gong. Spectrally Regularized Surfaces. PhD thesis, Diss. ETH No. 22616, MOSAIC Group, ETH Zurich, 2015. (PDF)

In order to ensure financial support for our project and allow further development of this software, please cite above publications in all your documents and manuscripts that made use of this software. Thanks a lot!


Image Naturalization

Image naturalization is an image enhancement technique that is based on gradient statistics of natural-scence images. The algorithm is completely parameter free. Simply open an image and choose "Enhancement - Naturalization" from the plugin menu. In fluorescence microscopy, image naturalization can be used for blind deconvolution, dehazing (removing scatter light), denoising, or contract enhancement. All just with one function! The "naturalness factor" displayed at the end tells you how close your original image was to a natural-scene one (1 meaning close, the farther from one the more different).

The algorithm implemented in this plug-in is described in:

Y. Gong and I. F. Sbalzarini. Image enhancement by gradient distribution specification. In Proc. ACCV, 12th Asian Conference on Computer Vision, Workshop on Emerging Topics in Image Enhancement and Restoration, pages w7–p3, Singapore, November 2014. (PDF)

In order to ensure financial support for our project and allow further development of this software, please cite above publications in all your documents and manuscripts that made use of this software. Thanks a lot!


Cluster processing support

If you have access to a parallel compute cluster, the MOSAICsuite can now automatically send and distribute jobs on the cluster and get back the results for you. Simply choose "Utility - Cluster" from the menu to create a profile for your compute cluster (machine name, login, etc.) and submit jobs there. You can even close your computer and go home while the jobs are running on the cluster. The next time to open Fiji again, it automatically displays the results or shows the current progress bar. No software installation on the cluster is necessary. Parallel computing has never been easier!


Region Competition Image Segmentation

This plugin can be used for multi-region image segmentation of 2D and 3D images without needing to know the number of regions beforehand. The method can handle objects with internal intensity gradients (i.e., shaded objects), but is limited to pixel-level accuracy and only delivers locally optimal results that depend on the initialization. See the user manual for details.

The algorithm implemented in this plug-in is described in:

J. Cardinale, G. Paul, and I. F. Sbalzarini. Discrete region competition for unknown numbers of connected regions. IEEE Trans. Image Process., 21(8):3531–3545, 2012. (PDF)

and in the Ph.D. thesis

J. Cardinale. Unsupervised Segmentation and Shape Posterior Estimation under Bayesian Image Models. PhD thesis, Diss. ETH No. 21026, MOSAIC Group, ETH Zurich, 2013. (PDF)

In order to ensure financial support for our project and allow further development of this software, please cite above publications in all your documents and manuscripts that made use of this software. Thanks a lot!


Split-Bregman Image Segmentation (a.k.a. the Squassh workflow)

The Squassh plugin can be used for globally optimal segmentation of piecewise constant regions in 2D and 3D images and for object-based co-localization analysis. Globally optimal segmentation is very fast and independent of initialization, but it is currently limited to objects with homogeneous (i.e., constant) internal intensity. The method can also produce sub-pixel accurate results. See the user manual for details.

The algorithm implemented in this plug-in is described in:

G. Paul, J. Cardinale, and I. F. Sbalzarini. Coupling image restoration and segmentation: A generalized linear model/Bregman perspective. Int. J. Comput. Vis., 2013. (PDF, Supplementary Material PDF, Matlab Software Download).

The Squassh workflow and protocol is described in:

A. Rizk, G. Paul, P. Incardona, M. Bugarski, M. Mansouri, A. Niemann, U. Ziegler, P. Berger, and I. F. Sbalzarini. Segmentation and quantification of subcellular structures in fluorescence microscopy images using Squassh. Nature Protocols, 9(3):586–596, 2014. (PDF, Supplementary Note PDF, Supplementary Data ZIP, Supplementary Video)

In order to ensure financial support for our project and allow further development of this software, please cite above publications in all your documents and manuscripts that made use of this software. Thanks a lot!


Spatial Pattern and Interaction Analysis

This plugin can be used for inferring spatial interactions between patterns of objects in images or between coordinates read from a file. See the user manual for details.

The theoretical framework and the algorithm implemented in this plugin are described in:

J. A. Helmuth, G. Paul, and I. F. Sbalzarini. Beyond co-localization: inferring spatial interactions between sub-cellular structures from microscopy images. BMC Bioinformatics, 11:372, 2010. (PDF)

and the plugin software itself in:

A. Shivanandan, A. Radenovic, and I. F. Sbalzarini. MosaicIA: an ImageJ/Fiji plugin for spatial pattern and interaction analysis. BMC Bioinformatics, 14:349, 2013. (PDF)

In order to ensure financial support for our project and allow further development of this software, please cite above publications in all your documents and manuscripts that made use of this software. Thanks a lot!


3D single-particle tracking

The evolution of our classical 2D single-particle tracking tool. The tool now also supports 3D images (but still includes support also for 2D ones) and includes new and improved algorithms. The documentation consists of: (1) The user manual of the original 2D version, (2) a tutorial with example data, and (3) an addendum describing the additional functionality in the 3D version. More information, including a source code API reference, can be found here: http://mosaic.mpi-cbg.de/ParticleTracker/.


Please use this version, even if you just plan to process 2D images. The old 2D plugin is no longer maintained.


November 2010: NEW VERSION!


The latest update of the 3D tracker plugin now supports:
- User-contributed function improvements (Many thanks to Dr. Kota Miura!)
- More robust detection algorithm
- Batch processing using ImageJ macros (output is written to disk)
- Number of detected particles is displayed
- Absolute threshold can be set instead of relative percentile
- The intensity momenta or order 0 to 4 are computed and output

The algorithm implemented in this plug-in is described in:

I. F. Sbalzarini and P. Koumoutsakos. Feature Point Tracking and Trajectory Analysis for Video Imaging in Cell Biology, Journal of Structural Biology 151(2):182-195, 2005. (PDF)

In order to ensure financial support for our project and allow further development of this software, please cite above publications in all your documents and manuscripts that made use of this software. Thanks a lot!


Background subtractor

This plugin implements a robust, histogram-based background subtractor, which is well suited to correct for inhomogeneous illumination artifacts. Please see the user manual for details and examples.


A PSF estimator

This plugin can be used to estimate the Point-Spread Function of the microscopy out of 2D fluorescence images. See the user manual for details.


3D PSF estimator

This plugin can be used to measure the 3D Point-Spread Function of a confocal microscope from an image stack. See the user manual for details.


Poisson noise

This plugin can be used to add synthetic Poisson-distributed noise to an image in order to simulate shot noise of various signal-to-noise ratios. It can be used to generate benchmark images in order to assess the accuracy and robustness of image processing algorithms as a function of the noise level present in images. See the user manual for details.


Bessel convolution

This plugin convolves an image with a Bessel function in order to simulate imaging with a microscope. The Bessel function is a model for a generic Point-Spread Function, depending on the wavelength and NA. This plugin can be used to generate synthetic images that mimic microscope output in order to benchmark the accuracy and robustness of image processing algorithms. See the user manual for details.


Spot detection

This small utility detects bright spots in images and estimates their center. It simply is the particle detection part of the particle tracker, without the linking. This can be handy in an image-processing pipeline or macro where only detection is needed. It can also be combined with Squash for segmentation patch positioning, and it handles time-series and video data.


The particle detection step is described in the following publication:

I. F. Sbalzarini and P. Koumoutsakos. Feature Point Tracking and Trajectory Analysis for Video Imaging in Cell Biology, Journal of Structural Biology 151(2):182-195, 2005. (PDF)

In order to ensure financial support for our project and allow further development of this software, please cite above publications in all your documents and manuscripts that made use of this software. Thanks a lot!


Region creator

A small utility to create manual segmentations to be used as ground truth to test and benchmark automatic segmentation algorithms.


Color substitution

A small utility to replace one color in an image with another color. While this is also possible by a combination of other ImageJ commands, this utility simplifies it to a one-step procedure.


IN NO EVENT SHALL THE MOSAIC GROUP BE LIABLE TO ANY PARTY FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN IF THE MOSAIC GROUP HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. THE MOSAIC GROUP SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS ON AN "AS IS" BASIS, AND THE MOSAIC GROUP HAS NO OBLIGATIONS TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.