calibcam
A command line tool for charuco-based calibration of multi-camera setups (intrinsic and extrinsic parameters), including omnidirectional cameras.
OpenCV, the popular computer vision library, provides tools for camera calibration, but they are not designed for multi-camera setups. calibcam fills this gap by providing a pipeline for multi-camera calibration.
Multi-camera calibration in calibcam works as follows:
First, OpenCV is used for single camera calibration, followed by an initial estimation of camera positions and orientations. Subsequently, all intrinsic and extrinsic parameters are optimised for reprojection error using Jax autograd.
Major features:
- Allows calibration of camera setups consisting of multiple lens types ex: a combination of omnidirectional and pinhole cameras.
- Flexibility to fix or optimize individual intrinsic and extrinsic parameters as needed, allowing users to leverage prior knowledge or constraints about their camera setup.
- Modular pipeline allows users to perform individual steps of the calibration process, such as detection, single camera calibration, and multi-camera calibration.
- Direct video support (.mp4 or .ccv) for calibration data input, given the videos are synchronised or have a constant frame offset.
- Generates plots in the output for visualisation of the calibration result, including spread of detections across cameras and reprojection error.
Note: See calibcamlib for a library for triangualtion, reprojection etc.
Board and data collection
calibcam uses Charuco boards for calibration (example board image below). See tools/board.py for the generation of both printable PNG and board configuration file for calibcam.
The board needs to be presented in different angles, positions and distances to each camera (important for accurate single camera calibration). Relative camera positions are estimated from frames in which the board is visible in multiple cameras. We recommend recording with 2 fps while moving the board around, spending around a minute on each camera.
Installation
Install bbo-calibcam via pip
pip install bbo-calibcam
or create conda environment from environment.yml:
conda env create -f https://raw.githubusercontent.com/bbo-lab/calibcam/main/environment.yml
Usage
- Collect data as described in Board section.
- Run calibcam with:
python -m calibcam --videos [LIST OF VIDEOS TO INCLUDE] --board [PATH TO BOARD.NPY file]
We recommend keeping a copy of the board file with the videos for documentation purposes.
- Check number of detections per camera in the output. Values should range between 80 and 300. If too few detections are made, check recording conditions (lighting, blur ...) and collect new calibration data. If too many frames are detected, convergence may be slow or run out of memory. Reduce detections adding a frame skip with
--frame_step. - Check reprojection error at the end of the output. Median errors should be <0.5px.
Format
Result
Generated multicam_calibration.npy/mat/yml holds a dictionary/struct with the calibration results. The filed "calibs" holds an array of calibration dictionarys/structs with entries
* 'rvec_cam': (3,) - Rotation vector of the respective cam (world->cam)
* 'tvec_cam': (3,) - Translation vector of the respective cam (world->cam)
* 'A': (3,3) - Camera matrix
* 'k': (5,) - Camera distortion coefficients
For further structure, refer to camcalibrator.build_result()
BBO internal MATLAB use only:
Use MATLAB function mcl = cameralib.helper.mcl_from_calibcam([PATH TO MAT FILE OUTPUT OF CALIBRATION]) from bboanlysis_m to generate an MCL file.
Release files for bbo-calibcam 4.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| bbo_calibcam-4.2.0.tar.gz | 41.6 kB | Without provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| bbo_calibcam-4.2.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 90.0 kB
Release files / bbo_calibcam-4.2.0.tar.gz
| Download URL | bbo_calibcam-4.2.0.tar.gz |
|---|---|
| Size | 41.6 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
945b884c1198758e3a306bbdc02be966fc1a6d8462e430d10b7349fd703efc84
|
|
BLAKE2b-256 checksum How to use checksums |
cb6efeb5e8a3352b0c5a82fef7ef2a330efff04e662f1708b09e114bde0d37e4
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
python-httpx/0.27.2
|
Release files / bbo_calibcam-4.2.0-py3-none-any.whl
| Download URL | bbo_calibcam-4.2.0-py3-none-any.whl |
|---|---|
| Size | 48.4 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
6af4a6e73aed0cb1235c5ff0a6d959b8486263f2c64113d4990448e8f9a64011
|
|
BLAKE2b-256 checksum How to use checksums |
f2920f1161becad42c444f84b838960bf4311d7a591daa0fe28c1c6f02562e98
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
python-httpx/0.27.2
|
Security for bbo-calibcam 4.2.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
