Upload checkerboard photos and calibrate a robot or vision camera without writing OpenCV code by hand.
Check reprojection error and per-frame quality before trusting a camera's calibration.
Export a calibrated camera model to ROS camera_info or Kalibr format for use in other software.
Run the whole calibration workflow locally instead of using the hosted version.
| musabali314/calibon | 000madz000/payload-test-api-route-handler | aliansari22/pdfclear | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | moderate | easy | easy |
| Complexity | 3/5 | 2/5 | 2/5 |
| Audience | developer | developer | general |
Figures from each repo's GitHub metadata at analysis time.
Local development needs both Node.js/npm and a Python 3.11+ virtual environment running side by side.
CalibON is a browser based tool for calibrating cameras, aimed at people working on robotics and computer vision projects. Camera calibration is the process of figuring out exactly how a camera distorts an image, so that measurements and 3D positions calculated from its photos can be corrected and trusted. CalibON walks a user through that process step by step instead of requiring them to write calibration code by hand. The workflow has five stages. First you set up the project by choosing a camera model, image resolution, and the size and layout of the checkerboard pattern used for calibration. Next you upload a set of photos of that checkerboard taken from different angles and distances. CalibON then detects the checkerboard corners in each photo using OpenCV, refines their positions, and shows quality metrics so you can accept or reject individual images. After that, it estimates the camera's internal properties and distortion using the accepted images. Finally, it shows the results, including how much error remains, per photo diagnostics, a preview of the corrected image, and files you can download. Under the hood, the frontend is built with React, TypeScript, and Vite, with Three.js used for the interactive visuals on the landing page. The backend is a Python FastAPI service that uses OpenCV and NumPy to do the actual calibration math. There is a hosted version at calibon.vercel.app that sends uploaded images to a backend for temporary processing, and the project also supports running both the frontend and backend on your own computer for local development, which requires Node.js, npm, and Python 3.11 or newer. Calibration results can be exported in several formats, including a general OpenCV JSON format, a ROS camera_info YAML file, and a Kalibr compatible file, so the calibrated camera model can be used in other robotics or vision software. Right now CalibON only supports the standard checkerboard pattern and a basic pinhole camera model. Support for other calibration patterns, fisheye lenses, webcam capture, and saving projects between browser sessions are all listed as planned future work. The project does not yet include a license file.
A browser based, step by step tool that calibrates a camera from checkerboard photos for robotics and computer vision use.
Mainly TypeScript. The stack also includes TypeScript, React, Vite.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.