Calibrate Subsets

SUMMARY

calibrate_subsets() calibrates and triangulates every camera subset of the given sizes, useful for finding which camera combination gives the most reliable rig.

SUPPORTED TARGETS

Available only on CharucoTarget, it calibrates and triangulates each subset, and both stages require per-corner ArUco ids.

UNITS

Input images are BGR ndarrays. Returned transforms are (4, 4) SE(3) matrices with translation in meters; reprojection errors are in pixels.

The Skill

subset_results = calibrator.calibrate_subsets(image_lists, subset_sizes=[2, 3], frame_index=0)

Skill	Returns	Description
calibrate_subsets(image_lists, subset_sizes=[2, 3], frame_index=0)	list[SubsetResult]	Calibrate and triangulate every camera subset of the given sizes.

Parameter	Type	Description
image_lists	list[list[ndarray]]	image_lists[cam][frame]; every camera must have the same frame count.
subset_sizes	list[int]	Subset sizes to evaluate. Default [2, 3].
frame_index	int	Frame to triangulate for each subset. Default 0.

The Code

from telekinesis.axon import MultiCameraCalibrator
from telekinesis.axon.targets import CharucoTarget

target = CharucoTarget(squares_x=6, squares_y=9, square_length=0.012, marker_length=0.009)

calibrator = MultiCameraCalibrator(target, num_cameras=4, reference_index=1)
subset_results = calibrator.calibrate_subsets(image_lists, subset_sizes=[2, 3], frame_index=0)

for subset in subset_results:
    subset.subset            # e.g. [0, 1]
    subset.result.ok
    subset.has_triangulation

SubsetResult

Field	Type	Description
subset	list[int]	Camera indices in this subset.
result	MultiCameraResult	Calibration result for this subset, see calibrate().
has_triangulation	bool	Whether triangulation was computed for this subset.
triangulation	TriangulationReport	Triangulation report for this subset, see triangulate().