Paired Plane Normal Error

Created by: LiDAR-LiDAR pairs

Overview

LiDAR points can't be directly compared to each other, but their alignment can be inferred. When MetriCal calibrates two lidars to one another, it uses the detected plane normals of the circle target from each LiDAR frame of reference to optimize. The difference in these normals is the Paired Plane Normal Error.

Note that the only LiDAR points used in this calculation are those detected on the retroreflective edge of the circle target; the interior points of the board are not used.

There is a Paired Plane Normal Error metric group for every pair of synced observations between LiDARs.

Description

Paired plane normal error metrics contain the following fields:

Field	Type	Description
from	UUID	The "from" component that these normal misalignments were computed in reference to.
to	UUID	The "to" component tha these normal misalignments were computed in reference to.
from_normals	An array of arrays of 3 floats	A collection of the XYZ orientations of plane normals in the "from" coordinate frame being matched against the to_normals.
to_normals	An array of arrays of 3 floats	A collection of the XYZ coordinates of plane normals in the "to" coordinate frame being matched against the from_normals.
misalignments	An array of arrays of 3 floats	The transformed chordal distance misalignment (spilt up according to the Cartesian / XYZ axes) between the to and from plane normals.
rmse	Float	The root-mean-square-error of all the misalignments.