When you link two units, your merged-cloud accuracy is roughly each camera’s own 3D accuracy plus the link calibration error. Note you can’t merge two clouds more accurately than each camera measures on its own. The link only adds error on top.
The error you can expect depends on your working distance and field of view. There is no universal “minimum” value. The exact values and tolerances of the error metrics depend on the measurement volume of your camera.
To verify, the result of the command EstimatePatternPose contains the error-metric values.
To achieve good calibration results:
- Mount both units rigidly, with overlapping fields of view on your region of interest, and keep them mechanically stable.
- Sort out synchronization before calibrating. Do not use software trigger for the link calibration. It can lead to poor calibration results within the link calibration procedure.
- For a well-conditioned link, collect pattern observations across the whole shared volume, not just one spot.
For more information, see Workspace Calibration and Measuring Calibration Accuracy.