.. _vision-hand-eye-calibration-example:

Hand-eye calibration
====================

Calibrate where a wrist camera sits on the arm (eye-in-hand) with the Tsai-Lenz
method, using the ChArUco board on the office world's table. The example moves
the wrist camera through 15 views of the board, varying its position and roll.
Each image is paired with the arm pose at its timestamp. The solver returns the
link-to-camera transform.

Use a UR e-Series or Franka arm at the office's default spawn point.

The result is printed, not persisted. Update the camera transform in your robot
description to use it. The same solver handles a fixed camera watching a board
on the arm (eye-to-hand), and
:meth:`~olo.vision.Calibration.calibrate_intrinsics` accepts the same ChArUco
frames when the camera itself needs calibrating.

.. tab-set::
    :sync-group: sdk-language

    .. tab-item:: Python
        :sync: python

        .. literalinclude:: scripts/vision-hand-eye-calibration.py
           :language: python

    .. tab-item:: TypeScript
        :sync: typescript

        .. literalinclude:: scripts/vision-hand-eye-calibration.ts
           :language: typescript