Installation¶
Again, two options are offered. One can first download a compiled version of
the Fiji plugin and try out B-recs in a minute. In that case, you simply
need to put the provided .jar
file in the Plugin folder of your
Fiji / Image J package. The second option is to compile yourself B-recs.
This is a more difficult task although scripts are provided to ease the
process. The major advantage of the second option is that you can then run
B-recs from the command line and the code is optimized for your system: it
allows optimal performances (see also code implementation)..
Requirements¶
A few standard libraries are required:
- cmake: Building tool for the package
- libbz2: each reconstructed image is stored as a raw image compressed with Bzip2 (when asked on the command line).
- libtiff: Brecs accepts input in tiff or raw format. All the results are output in the tiff format.
- fftw3: Fourier transforms are not used in the algorithms itself but only in the preprocessing of the datasets, in order to define regions of interests whereby the algorithm is applied.
- JNI: This is used in order to compile the java plugin.
- Sphinx: this is the format of the present documentation. You need this library to generate the documentation
Quick install¶
mkdir build
cd build
cmake .. -DCMAKE_C_FLAGS_RELEASE="-O3 -march=native -mtune=native"
make
make install
You should then be able to run brecs-reconssparse --version
. You can also
adapt the place where the program is install by adjusting the prefix:
cmake -DCMAKE_INSTALL_PREFIX:PATH=/install/folder ..
Mac OS X¶
A homebrew formula is provided in order to ease the process. Start by
installing Homebrew. Place the file packages/brecs.rb
from the B-recs
repository in the Formula folder, usually /usr/local/Library/Formula
Then, simply execute brew install brecs
.
You should then be able to run the command line version of B-recs
brecs-reconssparse
.
Development version¶
If you use the development version available on bitbucket, the install process is very similar. Refer to the cmake documentation for more details.