[Enhancement] Geolocation of ZED detections

[Kai-Shafe]

Why Is This Enhancement Needed?

Our vision models will be able to provide pixel coordinate values for the center of a detected obstacle / object. We need to calculate the UTM coordinate values of detections for use in pathfinding.

Proposed Solution

We will have to perform coordinate transformations based on the relative positions returned from the ZED cameras, then overwrite the coordinate values within the detection handlers.

Additional Context

Research the Active Object design pattern for more details on multi-threading considerations for the handlers.

[Brenn515]

We need to create a struct in object detector that gets the center point and radius. A good example is the Astar class in src/algorithms/planners/AStar. Important variables in that file are dNorthing which is the objects height for coordinate data, dEasting which is the objects width for coordinate data, and the struct at the bottom of the class that we should look at when doing our struct. It's also a good idea to look at the files in vision/aruco and src/drivers/NavigationBoard for how they do the vector points. Beyond that what we need to figure out is how to get the distance and the angle so we can get the measurements of the object seen by the rover.