type

states.VonMisesFisherUpdate

class VonMisesFisherUpdate:

Conjugate von Mises-Fisher update for matched tracklet-detection pairs.

For each matched pair the posterior is the resultant of the prior direction (weighted by its concentration) and the L2-normalised detection embedding (weighted by :attr:`kappa_obs`). The new direction is the normalised resultant and the new concentration is its length, so confident agreement sharpens the belief while disagreement broadens it. Unmatched tracklets keep their predicted state.

Parameters

NameTypeDescription
fieldstrTracklet field holding the mean direction.
meas_fieldstrDetection field holding the measured embedding.
kappa_obs = 20.0floatObservation concentration (how much one detection is trusted).

Members

method

property

Source: unitrack/states/directional.py:103