RFS-SLAM C++ Library Version 1.2

• Version 1.2 changes:
  • Implementation of joint compatibility branch and bound (JCBB) for data association in vector-based methods
  • Config files now use xml format to removed dependency on the libconfig library
  • OSX compatible when compiling with Clang/LLVM
  • Multi-threaded versions of SLAM algorithms using OpenMP
    • multithreading with OpenMP is currently not supported by Clang/LLVM
    • An OpenMP-supported LLVM compiler is available at: http://clang-omp.github.io/
  • Inclusion of the Victoria Park dataset and the code for processing it using various SLAM filters.
  • Performance profiling option using Google Perftools
  • Updates to CMakeLists.txt to make options more operating system specific
  • Executables now use the Boost program_options library to handle arguments

RFS-SLAM C++ Library v1.1

• RB-PHD SLAM algorithm updates:
  • included multi-feature particle importance weighting strategy
  • included single-cluster (SC)-PHD SLAM weighting strategy
  • new FastSLAM and MH-FastSLAM algorithms included
  • updated 2-D simulations
  • new Optimal Sub-Pattern Assignment (OSPA) error metric class
  • visualization tools are now in Python instead of Matlab
  • introduced namespace rfs to the library
  • some updates made to naming convention of classes in the library
  • cmake now generates rfsslam-config.cmake to enable find(rfsslam)
    from other projects