Library to work with calibration from bbo-calibcam
Install via pip as bbo-calibcamlib.
Load a calibration with
cs = calibcamlib.Camerasystem.load([PATH TO multicam_calibration.yml])Use calibrated camera system with
coords2d = cs.project(self, X, offsets=None)
# Project points in space of shape np.array((..., 3)) to all cameras.
# Returns image coordinates np.array((N_CAMS, ..., 2))
dirs, cam_pos = cs.get_camera_lines(self, x, offsets=None)
# Get camera lines corresponding to image coordinates in shape np.array((N_CAMS, ..., 2)) for all cameras.
# Returns directions from camera np.array((N_CAMS, ..., 3)) and camera positions np.array((N_CAMS, ..., 3))
# in world coordinates (for direct triangulation)
coords3d = cs.triangulate(self, x, offsets=None)
# Triangulate image coordinates in shape np.array((N_CAMS, ..., 2)) by minimizing reprojection error.
# Returns 3d points np.array((..., 3)) in world coordinates.
coords3d = cs.triangulate_3derr(self, x, offsets=None)
# Triangulate image coordinates in shape np.array((N_CAMS, ..., 2)) by finding the closest point to camera lines
# Returns 3d points np.array((..., 3)) in world coordinates.