blurrer

Module Contents

Classes

Blurrer

Simulate localization and vision noize.

class blurrer.Blurrer(object_angle_noize=0.0, object_distance_noize=0.0, observation_bonus=0.0, step_cost=0.0, constant_loc_noize=0.0, loc_noize_meters=0.0)[исходный код]

Simulate localization and vision noize. Params is placed in the blurrer.json file. :param object_angle_noize: Noize for angle in radians.

Blurrer will uniformly random value from -object_angle_noize to object_angle_noize and add it to the ground truth course. Defaults to 0.

Параметры
  • object_distance_noize (float, optional) – Noize for distance in percents divided by 100. Blurrer will uniformly random value from -object_distance_noize to object_distance_noize and multiply difference of 1 and this value with ground truth distance. Defaults to 0..

  • observation_bonus (float, optional) – Blurrer will increase the consistency for every good observation (successfuly processed image). Defaults to 0..

  • step_cost (float, optional) – Blurrer will decrease the consistency for every simulation step. Defaults to 0..

  • constant_loc_noize (float, optional) – Constant localization noize. Defaults to 0..

  • loc_noize_meters (float, optional) – Multiplier for consistency, in meters. Defaults to 0. Defaults to 0..

load_json(self, filename)[исходный код]
course(self, angle)[исходный код]
distance(self, distance)[исходный код]
objects(self, course=course, distance=distance)[исходный код]
loc(self, x, y)[исходный код]
coord(self, p)[исходный код]
step(self)[исходный код]
observation(self)[исходный код]
update_consistency(self, value)[исходный код]