• Linear algebra
  • Sparse and dense matrix data structures
  • Iterative linear solvers
    • CG
    • CGLS
  • Preconditioners
    • Jacobi
    • SSOR
• LAPACK wrappers
• Optimization
  • Levenberg-Marquardt
  • Split-Bregman
  • Reweighted least-squares
  • PEGASOS SVM solver
  • Kalman filter
• Tracking
  • Feature point data structure
  • Track administration
  • TST
• Geometry
  • Camera models
  • Coordinate transformations
  • B-splines in arbitrary dimensions
• Image descriptors
  • BRIEF
  • HOG
  • Raw image (with different normalizations)
  • Import/export
• Example applications
  • Robust estimation
  • CLAM (real-time structure from motion)
  • Multiview descriptor aggregation
  • TV image denoising
• Python bindings
  • File I/O
  • B-splines
• Documentation
• Unit testing

• OpenCV 3.0 (!)
• LAPACK
• QT (only required by examples with GUI)
• OpenMesh (required by parts of the reconstruction library)
• FFTW (optional)

An automatic build requires the following items:

• qmake
• Doxygen (documentation only)
• Latex (documentation only)
• GraphViz (documentation only)

Make sure that aforementioned dependencies are installed in default locations. In particular, it is highly recommended to build the latest OpenCV version (i.e., the trunk) from source. To build R4R, clone this repository by typing

git clone https://github.com/jonabalzer/r4r.git

This will automatically create a subfolder r4r. Change into this directory

cd r4r

and create a directory for the out-of-core build, say

mkdir build

Call qmake from the build directory:

cd build
qmake ..

If you want to include the example applications, you need to set the HAVE_EXAMPLES variable:

qmake .. "HAVE_EXAMPLES=1"

Start the build process by

make

The documentation is created via

make doc

If you plan to use the core libraries outside of Qt Creator, run

sudo make install

This will install libraries and header files into the appropriate system paths (which probably requires root privileges).