Electron tomography toolbox
eTomo is a python toolbox made for electron tomography
The basic forward and backward projection operations are based on the GPU-accelerated ASTRA-toolbox http://www.astra-toolbox.com/
This toolbox provides tools and reconstruction algorithms frequently used for building a electron tomographic pipeline, mainly for materials science.
- Numpy for data processing
- matplotlib for data visualization
- The astra-toolbox with GPU support. For installation of astra-toolbox, refer to http://www.astra-toolbox.com/
- Python environment: 64-bit 3.5 or 3.6
- Cuda 5.5 or hihger
- HypersPy: a library to process EDS/EELS spectroscopic images
- Operator Discretization Library (ODL): an operator-based package for optimization methods
TODO
Download the package, change to the source folder, type
pip install -e .
in the terminal
Support reading .mrc files. .mrc is a standard file format for tomographic measurement data from electron microscopes.
Basic functions are provided: Normalization: Alignment:
A slice viewer is provided basd on matplotlib. It is useful for viewing 3D volume / tilt-series of projection images Support viewing and ploting the reconstruction while the iterative reconstruction algorithm is running
Default reconstruction algorithms implemented as follows:
- Analytical algorithms: FBP (simple and fast)
- Numerical algorithms: SIRT (smooth reconstructions)
- Statistical algorithms: EM (for data with strong Poisson noise)
- Advanced primal-dual algorithms: Chambolle-Pock It is also possible to implement your own reconstruction algorithm
TODO
Examples
- TV-regularized STEM tomographic reconstruction
- Customized reconstruction algorithm
- HAADF-EDS bimodal tomographic reconstruction
- Nonlinearity correction for STEM tomography ...
This software is a product of the computational imaging group at CWI, Amsterdam https://www.cwi.nl/research/groups/computational-imaging
Copyright: 2018, CWI, Amsterdam http://www.cwi.nl/