-
Notifications
You must be signed in to change notification settings - Fork 18
Configuration Options
PLEASE NOTE: The information below is outdated. Configuration files for Burro are now located in the config/ module. The defaults are contained in config/defaults.ini.
Here's a brief explanation of all configuration options that reside in burro/config.py:
THROTTLE_CHANNEL = 2
The RC channel for Throttle
YAW_CHANNEL = 0
The RC channel for Yaw (i.e. steering)
ARM_CHANNEL = 4
The RC channel for arming (i.e. "activating" the vehicle")
THROTTLE_RECORD_LIMIT = 0.01
The throttle threshold to trigger recording (0.01 corresponds to 1%)
SESSION_DIR = "/home/pi/burro_sessions"
The directory to save recordings. Each session will be saved in a separate subdirectory.
CAMERA_RESOLUTION = (160, 120)
The resolution of the camera sensor
CAMERA_FRAMERATE = 30
The framerate of the camera. Note that the actual control loop may be actually limited by other factors, such as model computation time.
CAMERA_HORIZONTAL_FOV = 120.
The horizontal field of view of the camera sensor.
CAMERA_OUTPUT_RANGE = (0, 255)
The value range that is output for each channel of R, G, B from the camera.
REVERSE_STEERING = False
Use to reverse servo commands for steering
CAR_MAX_STEERING_ANGLE = 30.
Maximum (physical) steering angle of the car. Used for saving accurate steering values when recording.
DRIFT_GAIN = 0.15
Drift compensation at the output of the steering. This is the P (proportional) component of the controller.
LEFT_MOTOR_TERMINAL = 1
Terminal for the left wheel motor.
RIGHT_MOTOR_TERMINAL = 2
Terminal for the right wheel motor.
LEFT_MOTOR_MULT = 1.0
Throttle multiplier for the left motor.
RIGHT_MOTOR_MULT = 1.0
Throttle multiplier for the right motor.
LEFT_MOTOR_REVERSE = False
Reverse option for the left motor.
RIGHT_MOTOR_REVERSE = False
Reverse option for the right motor.
MODEL_INPUT_RANGE = (0, 1)
The range of inputs for each channel of R, G, B of the Keras model
MODEL_OUTPUT_SIZE = 15
The size (i.e. category count) of the model output
KERAS_AVERAGE_FACTOR = 0.8
This option controls the averaging factor b applied after the output of the Keras model.
Burro applies averaging to the Keras model output, to prevent excessive jitter at the output. Averaging is in the form w(t) := b*w(t) + (1-b)*w(t-1).