sighting_shows

Contents

sighting_shows#

rapidtools.processing.street_tracking.sighting_shows(obs: Observation, target: tuple[float, float], project: Callable[[float, float], tuple[float, float]], bearing_sigma_deg: float = 0.75, size_m: float = 5.0, range_ratio_bounds: tuple[float, float] = (0.6, 1.6)) → bool[source]#

Whether a sighting can be of an object at target.

The ray must pass within a quarter of the object’s size of the target (plus bearing noise) with the target in front of the camera, and the ground-contact range, when there is one, must agree with the distance to the target within range_ratio_bounds. The lateral tolerance is deliberately tighter than merge_by_rays() uses: a car parked alongside is 2.5 m across and must fail this test, while a piece of the object itself (half a car past a pole, its rear or front) is within about 1 m of the centre.

Parameters:

  • obs – The sighting (bearing set).

  • target – (x, y) in the local metre frame.

  • project – Local metre projection.

  • bearing_sigma_deg – One-sigma bearing error.

  • size_m – Object size the lateral tolerance scales with.

  • range_ratio_bounds – Accepted band for range over distance.

Returns:

bool – True when the sighting is consistent with the target.

Example

>>> from rapidtools.core import Observation
>>> from rapidtools.processing.street_localization import local_projection
>>> project, _ = local_projection(-117.4, 47.7)
>>> box = [(0.5, 0.5), (0.52, 0.5), (0.52, 0.55)]
>>> o = Observation('i', 'car', box, -117.4, 47.7, 0.0, bearing=90.0)
>>> o.range_m = 10.0
>>> sighting_shows(o, (10.0, 0.5), project)
True
>>> sighting_shows(o, (10.0, 3.0), project)  # 3 m across: the car alongside
False
Contents