D* Lite is an incremental heuristic search algorithm by Sven Koenig and Maxim Likhachev that builds on LPA*, an incremental heuristic search algorithm that combines ideas of A* and Dynamic SWSF-FP.
If you plan on building the code yourself, you will need the following:
- VC++ or GCC
- FLTK 1.3
bin/
Contains pre-built binary files to run on both Windows and Linux machines.
ide/
Contains the project files for Visual Studio 2010 (Windows) and Eclipse (Linux).
src/
Contains the source files of the project. The actual implementation of the D* Lite algorithm can be found in planner.h and planner.cpp (based on "Improved Fast Replanning for Robot Navigation in Unknown Terrain" by Sven Koenig and Maxim Likhachev - Figure 6: D* Lite: Final Version (optimized verion)).
The "maps" used for this project are grayscale bitmaps. The darker the pixel the more it costs to traverse it. Black pixels (RGB 0 0 0) are considered "unwalkable" while white pixels (RGB 255 255 255) cost "1". Pixel colors in between vary in value.
d-star-lite.exe "Simulator 1" "../maps/map-03-real.bmp" "../maps/map-03-robot.bmp" 0 0 249 249 30
To run the program, the following arguments must be passed via command line.
- [char*] Name of the window.
- [char*] Location of the real map.
- [char*] Location of the robot map.
- [int] Start x-coordinate.
- [int] Start y-coordinate.
- [int] Goal x-coordinate.
- [int] Goal y-coordinate.
- [int] Scanner radius.
Koenig, S and Likhachev, M. Improved Fast Replanning for Robot Navigation in Unknown Terrain. Atlanta : Georgia Institute of Technology, 2002.