Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/MartinHahner/LiDAR_fog_sim
LiDAR fog simulation
https://github.com/MartinHahner/LiDAR_fog_sim
3d autonomous-driving computer-vision fog fog-attenuation fog-removal lidar lidar-point-cloud point-cloud simulation
Last synced: about 2 months ago
JSON representation
LiDAR fog simulation
- Host: GitHub
- URL: https://github.com/MartinHahner/LiDAR_fog_sim
- Owner: MartinHahner
- License: other
- Created: 2021-08-11T12:23:04.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2022-06-28T03:52:24.000Z (over 2 years ago)
- Last Synced: 2024-08-01T03:44:37.354Z (5 months ago)
- Topics: 3d, autonomous-driving, computer-vision, fog, fog-attenuation, fog-removal, lidar, lidar-point-cloud, point-cloud, simulation
- Language: Python
- Homepage: https://www.trace.ethz.ch/lidar_fog_sim
- Size: 5.12 MB
- Stars: 169
- Watchers: 7
- Forks: 31
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
Created by [Martin Hahner](https://sites.google.com/view/martinhahner/home) at the [Computer Vision Lab](https://vision.ee.ethz.ch/) of [ETH Zurich](https://ethz.ch/).
[](https://arxiv.org/abs/2108.05249) [](https://paperswithcode.com/sota/3d-object-detection-on-stf?p=fog-simulation-on-real-lidar-point-clouds-for) 
# [Fog Simulation on Real LiDAR Point Clouds
for 3D Object Detection in Adverse Weather](https://arxiv.org/abs/2108.05249)
*by [Martin Hahner](https://www.trace.ethz.ch/team/members/martin.html), [Christos Sakaridis](https://www.trace.ethz.ch/team/members/christos.html), [Dengxin Dai](https://www.trace.ethz.ch/team/members/dengxin.html), and [Luc van Gool](https://www.trace.ethz.ch/team/members/luc.html)*
Accepted at [ICCV 2021](http://iccv2021.thecvf.com).
Please visit our [paper website](https://trace.ethz.ch/lidar_fog_sim) for more details.
## Overview
.
├── file_lists # contains file lists for pointcloud_viewer.py
│ └── ...
├── integral_lookup_tables # contains lookup tables to speed up the fog simulation
│ └── ...
├── extract_fog.py # to extract real fog noise* from the SeeingThroughFog dataset
├── fog_simulation.py # to augment a clear weather pointcloud with artificial fog (used during training)
├── generate_integral_lookup_table.py # to precompute the integral inside the fog equation
├── pointcloud_viewer.py # to visualize entire point clouds of different datasets with the option to augment fog into their scenes
├── README.md
└── theory.py # to visualize the theory behind a single LiDAR beam in foggy conditions
\* Contains returns not only from fog, but also from physical objects that are closeby.
**Datasets supported by `pointcloud_viewer.py`:**
- [H3D](https://usa.honda-ri.com/H3D)
- [A2D2](https://www.a2d2.audi/a2d2/en.html)
- [KITTI](http://www.cvlibs.net/datasets/kitti/eval_object.php?obj_benchmark=3d)
- [LyftL5](https://self-driving.lyft.com/level5/prediction/)
- [Pandaset](https://pandaset.org/)
- [nuScenes](https://www.nuscenes.org/nuscenes)
- [Argoverse](https://www.argoverse.org/data.html#tracking-link)
- [ApolloScape](http://apolloscape.auto/tracking.html)
- **[SeeingThroughFog](https://www.cs.princeton.edu/~fheide/AdverseWeatherFusion/)** :arrow_left: works best
- [WaymoOpenDataset](https://waymo.com/open/) (via [waymo_kitti_converter](https://github.com/caizhongang/waymo_kitti_converter))
## License
This software is made available for non-commercial use under a Creative Commons [License](LICENSE).
A summary of the license can be found [here](https://creativecommons.org/licenses/by-nc/4.0/).
## Acknowledgments
This work is supported by [Toyota](https://www.toyota-europe.com/) via the [TRACE](https://www.trace.ethz.ch/) project.
Furthermore, we would like to thank the authors of [SeeingThroughFog](https://www.cs.princeton.edu/~fheide/AdverseWeatherFusion/) for their great work.
In this repository, we use a [fork](https://github.com/MartinHahner/SeeingThroughFog) of [their original repository](https://github.com/princeton-computational-imaging/SeeingThroughFog) to visualize annotations and compare to their fog simulation. Their code is licensed via the [MIT License](https://github.com/princeton-computational-imaging/SeeingThroughFog/blob/master/LICENSE).
## Citation(s)
If you find this work useful, please consider citing our paper.
```bibtex
@inproceedings{HahnerICCV21,
author = {Hahner, Martin and Sakaridis, Christos and Dai, Dengxin and Van Gool, Luc},
title = {{Fog Simulation on Real LiDAR Point Clouds for 3D Object Detection in Adverse Weather}},
booktitle = {IEEE International Conference on Computer Vision (ICCV)},
year = {2021},
}
```
You may also want to check out our latest work (Oral at CVPR 2022)
[*LiDAR Snowfall Simulation for Robust 3D Object Detection*](https://github.com/SysCV/LiDAR_snow_sim).
```bibtex
@inproceedings{HahnerCVPR22,
author = {Hahner, Martin and Sakaridis, Christos and Bijelic, Mario and Heide, Felix and Yu, Fisher and Dai, Dengxin and Van Gool, Luc},
title = {{LiDAR Snowfall Simulation for Robust 3D Object Detection}},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2022},
}
```
## Getting Started
### Setup
1) Install [anaconda](https://docs.anaconda.com/anaconda/install/).
2) Create a new conda environment.
```bash
conda create --name foggy_lidar python=3.9 -y
```
3) Activate the newly created conda environment.
```bash
conda activate foggy_lidar
```
4) Install all necessary packages.
```bash
conda install matplotlib numpy opencv pandas plyfile pyopengl pyqt pyqtgraph quaternion scipy tqdm -c conda-forge -y
pip install pyquaternion
```
5) Clone this repository (including submodules).
```bash
git clone [email protected]:MartinHahner/LiDAR_fog_sim.git --recursive
cd LiDAR_fog_sim
```
### Usage
How to run the script that visualizes the theory behind a single LiDAR beam in foggy conditions:
```bash
python theory.py
```
How to run the script that visualizes entire point clouds of different datasets:
```bash
python pointcloud_viewer.py -d
```
**Note:**
You may also have to adjust the relative paths in `pointcloud_viewer.py` (right at the beginning of the file) to be compatible with your datasets relative folder structure.
### Disclaimer
The code has been successfully tested on
- Ubuntu 18.04.5 LTS
- macOS Big Sur 11.2.1
- Debian GNU/Linux 9.13
using conda 4.9.2.
## Contributions
Please feel free to suggest improvements to this repository.
We are always open to merge usefull pull request.