Unitree Robotics open-sources UNICALIB for robot camera calibration
Unitree Robotics open-sourced UNICALIB 0.1.7, a browser-based workbench that calibrates robot cameras and automatically checks whether the results pass. It supports intrinsic, extrinsic and hand-eye calibration on x86-64 Linux. Live capture requires TeleImager commit 940fd69 and does not support v2.1.0 or later.
China context
- Original name
- 宇树科技
- Outside China
- Not stated in the sources yet
- Claims
- Company-reported; not yet independently evaluated
- For builders
- The GitHub repository documents an English interface and a custom-data workflow that does not need a live robot connection. The source does not establish whether the documented live setup works on specific overseas robot configurations; verify the robot model and installed TeleImager version before planning a live calibration.
Unitree Robotics published UNICALIB on GitHub. The workbench guides users through calibration-board poses, selects images, runs the calibration and can copy the results to the robot. It also accepts existing images as custom data without requiring the robot's live image service.
UNICALIB supports pinhole and Kannala–Brandt fisheye camera models. The README lists live hand-eye calibration for the G1 D, 29-DoF G1 and R1-A7-D, with model-specific setup constraints. A useful technical check is whether repeated calibrations on the same robot produce consistent results across operators.
Unitree robot integrators can use one browser interface to capture calibration images, check the results and write them back to the robot. Teams running TeleImager v2.1.0 or later cannot use live capture with this version of UNICALIB; they can instead calibrate from existing images. The next compatibility signal is a release note confirming support for the newer TeleImager camera configuration.
For teams maintaining Unitree robots, the practical question is whether guided capture and automatic checks reduce technician time or repeat calibration. The repository does not quantify either outcome; a side-by-side maintenance trial could measure both.
Check future repository releases for support for newer TeleImager versions and additional robot configurations. Independent tests that report setup time and calibration repeatability would show how the workflow compares with existing tools.