Skip to content

Latest commit

 

History

History

sample_code_cpp

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
data
data
 
 
images
images
 
 
src
src
 
 
README.md
README.md
 
 
framework.cfg
framework.cfg
 
 

README.md

Sample SLAM

Scan matching as a front-end, Graph based SLAM as a back-end. Inspired by furo-org/LittleSLAM.

Odometry only Sample SLAM
odom.png slam.png

The framework can be customized in framework.cfg by Factory Method as shown below:

Module Type Description
PointCloudMap _EntirePoint Stores entire laser point
_GridTable Stores laser points in a grid based manner
_SubMap Stores laser points in a grid based manner with sub maps
ReferenceScanMaker _LastScan Last scan as a reference scan for scan matching
_LocalMap Local map as a reference scan for scan matching
PoseEstimator _ICP Iterative Closest Point
_NDT Normal Distribution Transform (WIP)
DataAssociator _LinearSearch Linear search to find corresponding laser points
_GridTabel Grid based search to find corresponding laser points
PoseOptimizer _GradientDescent Gradient descent with a fixed step rate
_LineSearch Gradient descent with a dynamic step rate determined by line search
CostFunction _EuclideanDistance Euclidean distance as an error distance
_PerpendicularDistance Perpendicular distance as an error distance

dependency_graph.png

Pre-requirements

  • Boost
  • CMake
  • Eigen
  • gnuplot

Install on Ubuntu

$ sudo apt-get install build-essential libboost-all-dev cmake libeigen3-dev gnuplot gnuplot-x11

Install on Mac

$ brew install boost cmake eigen gnuplot

How to build

$ git clone --recursive https://github.com/hotsuyuki/Graph-Based-SLAM.git
$ cd Graph-Based-SLAM/
$ cd sample_code_cpp/src/
$ mkdir build
$ cd build/
$ cmake ..
$ cmake --build .

This command downloads google/googletest using Crascit/DownloadProject and produces an executable file named main in the build directory.

How to run

The executable file main requires at least two arguments:

  • data-file ... path to the input scan data
  • viz-mode ... visualization mode: scan | odom | slam

(a) Launch SLAM

$ ./main --data-file ../../data/corridor.lsc --viz-mode slam

or

$ ./main ../../data/corridor.lsc slam

(b) Odometry only

$ ./main --data-file ../../data/corridor.lsc --viz-mode odom

or

$ ./main ../../data/corridor.lsc odom

(c) Check scan data

$ ./main --data-file ../../data/corridor.lsc --viz-mode scan

or

$ ./main ../../data/corridor.lsc scan

The full arguments and options are shown below:

usage:
  main [OPTION...] data-file viz-mode

positional arguments:
  data-file  String: Path to sensor data file (.lsc)
  viz-mode   String: Visualization mode ... scan | odom | slam

options:
  -h, --help  Print usage

How to test

The program can be tested with Google C++ Testing Framework.

$ cd Graph-Based-SLAM/
$ cd sample_code_cpp/src/build/
$ ./test/utility/utility_test

References