FAIR: Flexible Algorithms for Image Registration - Software and Apps
- What is it? FAIR stands for Flexible Algorithms for Image Registration and is a package written in MATLAB which can be used for solving the so-called correspondence problem. The objective is to automatically establish point-to-point correspondences between objects in two different scenes; see [FAIR 2009] for more information. The package comes in combination with a PDF of [FAIR 2009] and a folder full of examples.
- Purpose? FAIR is primarily designed as an academic and teaching tool. It can be used to explore existing techniques or to invent new features (please let us know). Though the focus is on exploring methods, FAIR can be - and already has been - used as a prototyping tool for practically relevant registration problems.
- What is new? A major feature of the current 2011 version is a novel nonlinear hyperelastic regularization energy. This regularizer allows for very large deformations which at the same time are very smooth and in particular diffeomorphic. Furthermore, the package is supplemented by parallelized C versions of computationally costly functions such as interpolation and distance measures. We also enabled the integration of individual add-ons which allows the fine tuning and individualization of the toolbox for specific applications.
- How to get it? To receive the FAIR toolbox, please download, complete, and sign the copyright form and send it either by email to email@example.com or by fax to +1-215-386-7999. You will be mailed a ZIP archive containing all m-files as well as a PDF version of the book. Please save your copyright form as a TXT, JPG, or PDF file.
- Demos and Screenshots The following list presents all intermediate results and the m-file for a non-parametric multi-level image registration with an affine linear pre-registration for 2D PET-CT data using the normalized gradient field as a distance measure and elastic regularization; a Gauss-Newton scheme is used as optimizer in a matrix free setting (using multigrid for the linear solve). The time was about 10sec for 128x128 data on a standard PC including all graphics and data pre-processing. [mfile] [data] [multi-level representation] [PIR results level 4] [NPIR results level 4] [NPIR results level 5] [NPIR results level 6] [NPIR results level 7] [final results] [output]
Available AppsFAIR not only provides opportunities to develop specific building blocks but also to redesign fundamental structures. As modifications can be severe, we decided to split the toolbox into a kernel and add-ons. The stand alone kernel provides all the basic functionality. The add-ons can be tuned to even overlay standard behaviour. In the following, we present some examples. We hope to extend this list with the help of your contributions.
- EPI please contact Fabian Gigengack or Lars Ruthotto
- DTI Correction please contact Fabian Gigengack or Lars Ruthotto
- VAMPIRE please contact Fabian Gigengack or Lars Ruthotto
To order this book click here.