Add Board-related logic to Pose Solver

[vaughantnrc]

Related to #6

[tavaughan]

@Rampex1 This issue refers to code that is still on NRC's gitlabc. Right now the Pose Solver only tracks single markers. If I remember correctly, the strategy for tracking a board would look something like this:

1. If multiple detectors see the target, build a geometry based on point visibility.
   • If multiple detectors see the same points (markers) and the reference, use intersection of rays as points
   • If only one detector sees points (markers) and the reference, represent the rays as lines
   • Then use an iterative closest point algorithm (combination of point-to-point and point-to-line) to determine the transform from the target to the reference
2. If only one detector sees the target, use solvePnP to determine transform from target to detector

The logic exists in code on gitlabc. I can migrate it later. I think it makes sense to unassign yourself from this issue.

[vaughantnrc]

80 shall theoretically address the missing logic in pose solver, though we still lack means of sending the target board to the pose solver.

[vaughantnrc]

Issue for sending the target board to the pose solver: https://github.com/PerkLab/MCSTrack/issues/84