Fence margin direction #31009
Fence margin direction #31009
Conversation
andyp1per
force-pushed
the
pr-fence-direction
branch
4 times, most recently
from
4ee5ef8 to
a228884
Compare
andyp1per
force-pushed
the
pr-fence-direction
branch
from
a228884 to
abea932
Compare
|
this feature looks like a good step towards resolving the long-term issue of RTL not working if the vehicle starts outside the fence. The only concern is around the Polygon_closest_distance_point() method and whether it requires more CPU especially on fences with lots of points |
|
It's not a breach is it? This seems to be calculating the distance to the nearest fence from the current location. There is no breach. I think it needs a different name maybe get_nearest_direction_NED()? |
Yes this method is probably O(n^2), it's polynomial for sure, which means it won't scale. This is why the OA library does these calculations in a kind of background thread. The code is also fundamentally the same as the code in the OA library. Should this not be a call into the existing OA library code? |
andyp1per
force-pushed
the
pr-fence-direction
branch
2 times, most recently
from
fd1c1f0 to
07f9e78
Compare
|
needs to pass CI :-) |
andyp1per
force-pushed
the
pr-fence-direction
branch
from
07f9e78 to
c1ef3ce
Compare
andyp1per
force-pushed
the
pr-fence-direction
branch
from
c1ef3ce to
826b609
Compare
… Polygon_closest_distance_point()
add get_breach_direction_NED() to return direction to closest breach
andyp1per
force-pushed
the
pr-fence-direction
branch
from
329041d to
88d7048
Compare
optionally return last breach position
andyp1per
force-pushed
the
pr-fence-direction
branch
from
9807390 to
ef0431e
Compare
5 participants
This provides a new function get_breach_direction_NED() which gives the direction of the closest fence matching the provided bitmask, akin to what get_breach_distance() does.