https://github.com/henriktrom/multi-camera-calib
Package for calibrating synchronized multi-camera setups, designed for robotic platforms. It captures calibration data, runs ChArUco-based calibration, and validates the results with 3D backprojection.
https://github.com/henriktrom/multi-camera-calib
camera-calibration multi-camera-calibration open-source research-software
Last synced: 10 days ago
JSON representation
Package for calibrating synchronized multi-camera setups, designed for robotic platforms. It captures calibration data, runs ChArUco-based calibration, and validates the results with 3D backprojection.
- Host: GitHub
- URL: https://github.com/henriktrom/multi-camera-calib
- Owner: HenrikTrom
- License: cc0-1.0
- Created: 2025-05-28T12:18:31.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-11-25T13:55:37.000Z (6 months ago)
- Last Synced: 2025-11-28T19:40:14.149Z (6 months ago)
- Topics: camera-calibration, multi-camera-calibration, open-source, research-software
- Language: C++
- Homepage:
- Size: 297 KB
- Stars: 1
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
- Citation: Citation.cff
Awesome Lists containing this project
README
# đź“· multi-camera-calib
[](https://zenodo.org/badge/latestdoi/991966353)
**multi-camera-calib** is a package for calibrating synchronized multi-camera setups, designed for robotic platforms. It captures calibration data, runs ChArUco-based calibration, and validates the results with 3D backprojection.
---
## đź“‘ Citation
If you use this software in your work, please cite it using the **“Cite this repository”** button on the top-right of this page.
## 🔧 Installation
### Dependencies
This package forms an essential part for my [docker-based real-time 3d tracker](https://github.com/HenrikTrom/ROSTrack-RT-3D). It mainly depends on these two modules
* [cpp_utils](https://github.com/HenrikTrom/cpp_utils)
* [flirmulticamera](https://github.com/HenrikTrom/flirmulticamera)
To build the package, simply run:
```bash
./scripts/build.sh
```
## Usage
### 1. Prepare a Calibration Target
Use a ChArUco board similar to the one shown in the example image above.
To get you started we provide a bitmap file in `./content/board_pattern.MCC_Patt1040x720.bmp`. You can use [GIMP](https://www.gimp.org/) to rescale and print the file with a size of 1040 x 720 mm, including the white border. The configuration bellow matches this exact calibration target.
### 2. Configure Calibration Settings
Edit the configuration file `./cfg/CameraCalibrationSettings.json` to match your setup:
```json
{
"pattern_size_first": 11,
"pattern_size_second": 7,
"charuco_params_border_size": 1.0, # pixels
"charuco_params_dict": 0, # dict id
"charuco_params_marker_border_pixel": 1,
"charuco_params_marker_to_square_ratio": 0.75,
"charuco_params_square_size": 80.0, # mm
"savedir": "/home/docker/workspace/workspace/multi-camera-calib/data/results"
}
```
### 3. Run the calibration routine
Start the calibration process:
```bash
./scripts/calibration.sh
```
#### What the script does:
* Waits 10 seconds for you to get the calibration board into view
* Records video for 15 seconds from each camera
* Extracts frames from the video streams
* Runs the calibration using the extracted frames
* Validates the calibration and stores results in `./test`
Make sure to move the calibration board smoothly and visibly during the recording phase (approx. 25 seconds total).
## Acknowledgements
The core of the calibration code is adapted from [Multi\_Camera\_Calibration](https://gitlab.com/ungetym/Multi_Camera_Calibration/-/blob/master/LICENSE?ref_type=heads). To streamline re-calibration, the GUI was removed and replaced with JSON-based configuration.