2D MCL with mixture proposal distributions
This library is used to perform kernel density estimation and kdtree sampling.
$ sudo apt-get install gcc g++ make pkg-config python libgsl0-dev libblas-dev liblapack-dev libboost-all-dev
$ git clone https://github.com/chungying/nuklei.git
$ cd nuklei
$ ./scons.py
$ sudo ./scons.py install
This library is for kdtree and 1-nearest-neighbor search.
$ sudo apt-get install -y libflann-dev
Most packages could be installed via rosdep or apt-get. Following two packages should be installed as well.
- MCL with mixture proposals by using pre-built sampling
- MCL with mixture proposals by using Evolutionary Markov Chain Monte Carlo sampling
Some results for AMCL, MCL, MIXMCL, and MCMCL (the proposed method) are shown here.
mp: maximum particle number
ri: resample intervel
