Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/jmuwrobotics/libbicos
GPU-Accelerated Binary Correspondence Search for Multishot Stereo Vision
https://github.com/jmuwrobotics/libbicos
computer-vision cuda depth-map stereo-camera stereo-matching stereo-vision
Last synced: 4 days ago
JSON representation
GPU-Accelerated Binary Correspondence Search for Multishot Stereo Vision
- Host: GitHub
- URL: https://github.com/jmuwrobotics/libbicos
- Owner: JMUWRobotics
- License: lgpl-3.0
- Created: 2024-09-17T09:30:58.000Z (4 months ago)
- Default Branch: main
- Last Pushed: 2024-12-17T08:36:16.000Z (17 days ago)
- Last Synced: 2024-12-30T05:11:57.340Z (4 days ago)
- Topics: computer-vision, cuda, depth-map, stereo-camera, stereo-matching, stereo-vision
- Language: Cuda
- Homepage:
- Size: 342 MB
- Stars: 0
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: COPYING
Awesome Lists containing this project
README
# libBICOS
BInary COrrespondence Search for multi-shot stereo imaging, with optional CUDA acceleration.
## Citing:
This is the implementation of the [corresponding paper](https://isprs-archives.copernicus.org/articles/XLVIII-2-W7-2024/57/2024/isprs-archives-XLVIII-2-W7-2024-57-2024.pdf) which appeared in [Optical 3D Metrology 2024](https://o3dm.fbk.eu):
```bibtex
@article{liebender2024libbicos,
title={libBICOS -- An Open Source GPU-Accelerated Library implementing BInary COrrespondence Search for 3D Reconstruction},
author={Liebender, Christoph and Bleier, Michael and N{\"u}chter, Andreas},
journal={The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences},
volume={48},
pages={57--64},
year={2024},
publisher={Copernicus Publications G{\"o}ttingen, Germany}
}
```
## Build:
Dependencies:
- `gcc` or equivalent C compiler with C++17 support (build)
- [`meson`](https://github.com/mesonbuild/meson) >= 1.1.0 (build)
- `opencv 4.x` with cuda support
- `cuda 12.x` including toolkit
- [`benchmark`](https://github.com/google/benchmark) (optional, for executing benchmarks)
```bash
# recommended: clone the most recent release
$ git clone --depth 1 --branch v1.2.2 https://github.com/JMUWRobotics/libBICOS
$ cd libBICOS
$ meson setup builddir --buildtype release
# optional but recommended if you have access to a CUDA-capable GPU
$ meson configure -Dimplementation=cuda builddir
$ meson install -C builddir
```
The versioning scheme of the library is [Semantic Versioning](https://semver.org/).
## Usage
### Linking
After installing, you can include `/usr/local/include/BICOS/*` and link against `/usr/local/lib/libBICOS.so`.
### Commandline-interface
Alternatively, this project builds `bicos-cli`. To give you a feel for the parameters of BICOS, you can download an example dataset with [`data/prepare.sh`](/data/prepare.sh) that contains rectified imagery from a stereo camera, where images per side only differ in the projected light pattern.
Calling:
```console
$ bicos-cli data/{left,right} -q data/Q.yaml --threshold 0.96 --stacksize 33 --limited --variance 2.0 --step 0.1 -o /tmp/result.png
```
will get you:
#### Disparity
#### Pointcloud
While only requiring ~110ms (RTX4090) and ~1.6GB VRAM for matching on two stacks of 33 images each.
The most significant parameters can be summarized:
- `--threshold`: increasing this reduces match-outliers. High values rely on a reliable pattern.
- `--variance`: high values reduce coarse outliers, e.g. pixels where no pattern was projected. May reduce correct matches on low pattern contrast.
- `--step`: optional value for subpixel-interpolation around matches.
Other settings are available; Try `bicos-cli -h` for details.
## Benchmarking:
```console
$ meson test --benchmark -C builddir --interactive
```
## Light projection:
For starters, you may find https://github.com/Sir-Photch/VRRTest useful for projecting a light pattern using an aftermarket projector.
## License
This library is licensed under the GNU Lesser General Public License v3.0 (or later).
Please see [COPYING](/COPYING) and [COPYING.LESSER](/COPYING.LESSER) for details.
