GPisMap_v2

This repository contains source codes and demo files for our paper Online Continuous Mapping using Gaussian Process Implicit Surfaces (GPIS), which is presented at IEEE ICRA 2019.

The previous repo GPisMAP is not maintained but kept as legacy. This repo contains some updates including slightly enhanced speed and simple python interfaces.

License

Licensed under GNU General Public License version 3.

Requirements

Eigen is the only external library required by GPIS itself.

• Ubuntu: sudo apt install libeigen3-dev cmake build-essential
• macOS: brew install cmake eigen@3

For running the demos and visualizing results, choose one of the two options below:

1. (Option 1) Build the source via MATLAB mex and run demo & visualization scripts on MATLAB. This is recommended for best visualization.

2. (Option 2) Alternatively, build the source via cmake and run demo & visualization scripts in Python (tested with Python 3.14 and CMake 3.20+). Visualization is limited and there are dependencies to be installed, but useful in case you have no Matlab license.

Compiling and Running

1. Clone this repository

git clone https://github.com/leebhoram/GPisMap2.git
cd GPisMap2

MATLAB Option

2. cd to the mex directory in MATLAB

cd <GPisMap2>/matlab/mex

3. Compile the mex functions by executing the make script.

   • Setup mex

   mex -setup
   

   • Run the make scripts

   make_GPisMap
   make_GPisMap3
   

4. Run the demo scripts

First,

cd <GPisMap2>/matlab

Then,

• For 2D

run('demo_gpisMap.m')

• For 3D

run('demo_gpisMap3.m')

5. Troubleshooting
   • If mex complains about not finding eigen, configure the eigen path appropriately in both make_GPisMap.m and make_GPisMap3.m

Python Option

2. Build the source

mkdir build && cd build
cmake ..
make -j $(nproc)              # Linux
make -j $(sysctl -n hw.ncpu)  # macOS

3. Install Python dependencies (Python 3.14). Use any environment manager you prefer — venv, conda, pyenv, etc. For example, with venv:

python3.14 -m venv .venv
source .venv/bin/activate

Then install:

pip install -r requirements.txt

If trying with PyVista (note that its dependency vtk takes ~600MB disk space):

pip install pyvista>=0.44

4. Run the demo scripts

cd <GPisMap2>/python

• For 2D

python test.py

• For 3D

python test3d.py                       # default using matplotlib (no alpha)
python test3d.py --pyvista --alpha 0.3 # if using pyvista (with alpha value)

Author

Bhoram Lee E-mail: <first_name>.<last_name>@gmail.com

Misc.

The main code was developed on Ubuntu 22.04, tested on Ubuntu 22.04 and macOS 15.5 (Apple Silicon).

Citation

If you find GPisMap/GPisMap_v2 useful in your research, please consider citing:

  @article{<blee-icra19>,
      Author = {Bhoram Lee, Clark Zhang, Zonghao Huang, and Daniel D. Lee},
      Title = {Online Continuous Mapping using Gaussian Process Implicit Surfaces},
      Journal = {IEEE ICRA},
      Year = {2019}
   }