Copter: support MAV_FRAME_LOCAL_NED in MISSION_ITEM for auto mode

[amilcarlucas]

This is based on #8748 and #15854

Here is the original description from @chobitsfan :

For indoor flight with external navigation data (ATT_POS_MOCAP or VISION_POSITION_ESTIMATE), it is not possible to flight auto mission with latitude/longitude waypoint.

This PR allows mission waypoint be assigned in MAV_FRAME_LOCAL_NED coordinate system. To do this, GCS can send mission_item message with frame = MAV_FRAME_LOCAL_NED (1).

Currently only takeoff and waypoint support local_ned frame

Test flight video, this mission contains 4 command: takeoff, waypoint 1 (local coordinate 1.2, 1.2) , waypoint 2 (local coordinate 1.2,-1.2) and land
2 waypoints are marked in red cross on the floor
I use mavproxy wp load for uploading mission to copter
https://youtu.be/gVgK41gS7Hc

And some more testing was done by @roackb2 :

Here's the video
This is the flight plan
And this is the parameters I use
Hardware: Pixhawk 2.4
Vision: Intel Realsense T265
Companion Computer: Raspberry Pi 3B+
script running on companion computer to feed position data to flight control is here

[amilcarlucas]

All CI tests pass now :)

[tridge]

this changes the binary format of our mission storage. I am concerned that this could break existing missions in storage on stm32 or linux boards to re-interpret

[rmackay9]

We discussed this at length on the dev call but had these concerns:

• the changes to the eeprom format for missions means we need careful testing on various boards to ensure following bits aren't affected
• relative yaw is not handled
• although it is huge scope creep ideally we would move the mission storage to a variable size format instead of the fixed format we use now. This would allow more waypoints (not that anyone is complaining) but perhaps also allow more flexibility in changing the size and/or format.

[amilcarlucas]

@chobitsfan can you follow up on this one? This is out-of-my-league

[chobitsfan]

Hi @amilcarlucas Yes. I will try to do it

[chobitsfan]

Hi @amilcarlucas I created #16627. try to fix the problem raised by @tridge in #16259 (comment)

[amilcarlucas]

@chobitsfan Thanks for the follow up