reminder.txt

#== Idea ==#
Packages to provide detection, localization, and mapping in the ROS environment by utilizing OpenCV/ARUCO.

#== Nodes ==#
ml_detector:
	- Use the in-built libraries to identify markers
	- Loads marker config from yaml file
	- Use yaml to switch between single and board (board is an array)
	- Boards are defined by:
		- Array of markers (split into height/width based on array)
		- Marker size
		- Marker separation
	- Singles or boards are treated as singular markers
	- Singles and boards cannot be used at the same time
	- Output:
		- Use marker_msgs/MarkerStamped (1 message per "marker")
		- ids[] lists all ids in a marker
		- id_confidence lists which singles could be seen in the marker (0 or 1)
		- pose is from marker center

ml_landmarks:
	- How mapping works:
		- Notes down first marker pose as the 0,0 point
		- Matches multiple markers with the same stamp as being linked
		- Link each marker to its parent, and back to 0,0
		- Push to TF, and perhaps filter future frames to refine transform
	- Should publish the camera frame relative to one of the markers (perhaps try to hold the highest in the TF tree)