Ray Casting & Diffusion (RCD) model for global path planning

Overview

RCD is a fast and efficient sampling-based global path planner that given a binary occupancy grid, the target's and the robot's position produces a collision free path.

Build

Get the repo:

git clone https://github.com/MichaelMarav/rcd_path_planner.git

Go to the rcd_path_planner/ folder

mkdir build && cd build

cmake .. -DCMAKE_BUILD_TYPE=Release

make -j8

Interactive grid plotter

In order to generate your own custom occupancy grid, install Pillow:

pip install Pillow

and then run the script "scripts/interactive_grid.py"

python3 interactive_grid.py

a set of instructions will guide you to easily design your experiment.

Run RCD

Modify the config file ("config/rcd_params.yaml") to specify which file you want to run RCD on.

Then go to the build folder and run RCD:

./rcd