repository·master·Indexed 43 days ago
https://github.com/amymcgovern/pyparrotA Python library providing an interface to control Parrot Mambo and Bebop 2 drones, designed for autonomous flight programming and educational STEM applications. It supports WiFi and Bluetooth Low Energy (BLE) connections, offering functionality for flight control (takeoff, landing, direct and relative movement), camera management, and real-time sensor data handling via callbacks.
The pyparrot package is organized into several subpackages and modules designed for drone control, vision, and networking.
pyparrot.commandsandsensors: Likely contains interfaces for drone commands and sensor data handling.pyparrot.networking: Handles network communication protocols.pyparrot.scripts: Contains executable scripts or utility scripts.pyparrot.utils: General purpose utility functions.pyparrot.Bebop: Interface for controlling Parrot Bebop drones.pyparrot.DroneVision: Provides drone vision capabilities.pyparrot.DroneVisionGUI: Graphical User Interface for drone vision tasks.pyparrot.Minidrone: Interface for controlling Minidrone hardware.pyparrot.VisionServer: Provides server-side capabilities for vision data streaming or processing.The pyparrot.networking package provides submodules for establishing network connections with drones. It supports two primary communication methods:
pyparrot.networking.bleConnection module.pyparrot.networking.wifiConnection module.Developers should use these modules to abstract the underlying transport layer when sending commands or receiving telemetry from a drone.
When using BLE, you must NEVER use time.sleep(), as it will cause the BLE connection to disconnect regularly. Instead, use smart_sleep(seconds) (where seconds can be a float). This method sleeps while remaining awake for BLE or WiFi notifications. If you are using WiFi, time.sleep() is acceptable, but smart_sleep() is compatible with both.
# Correct way to pause in a BLE-connected script
import pyparrot
mambo.smart_sleep(2.5)The required interface depends on your specific Parrot drone model:
| Drone Model | Interface | OS Support |
|---|---|---|
| Parrot Mambo FPV | WiFi | Mac, Linux, Windows |
| Parrot Mambo Fly/Code | BLE | Linux only |
| Parrot Swing | BLE | Linux only |
| Parrot Bebop 2 | WiFi | Any WiFi enabled device |
| Parrot Bebop 1 | WiFi | Any WiFi enabled device |
| Parrot Anafi | SDK (Pending) | TBD |
The DroneVisionGUI class provides a vision approach using the libVLC library. This method processes video entirely in memory, minimizing delay compared to the ffmpeg approach. It is ideal for real-time vision tasks where you do not need to save the full image stream to disk.
Requirement: You MUST install the VLC client version 3.0.1 or greater. Installing only libvlc is insufficient; the full VLC client installs the extra software required by the libvlc Python interface.
To control the drone while vision is running, you must provide a function to the user_code_to_run parameter in the DroneVisionGUI constructor.
mamboVision = DroneVisionGUI(mambo, is_bebop=False, buffer_size=200,
user_code_to_run=demo_mambo_user_vision_function,
user_args=(mambo, ))Sensor data is stored in a BebopSensors object. The most efficient way to consume sensor updates is to register a callback function that triggers every time the sensors are updated (typically at 10Hz via WiFi).
Use set_user_sensor_callback(function, args) to register your handler.
battery: Battery level (defaults to 100 until real data arrives).flying_state: Current state. Possible values: "landed", "takingoff", "hovering", "flying", "landing", "emergency", "usertakeoff", "motor_ramping", "emergency_landing".sensors_dict: A dictionary containing all other sensor data defined in the drone's XML configuration.def my_sensor_handler(sensor_data, user_arg):
print(f"Drone state: {sensor_data.flying_state}")
print(f"Battery: {sensor_data.battery}%")
# Register the callback
drone.set_user_sensor_callback(my_sensor_handler, "my_custom_arg")pyparrot.networking.wifiConnection module to manage connections to drones over a Wi-Fi network. This module provides the interface for establishing and managing socket-based or protocol-specific communication over Wi-Fi.To avoid unreliable RTSP/RTP streams in OpenCV, pyparrot provides a vision approach using ffmpeg. This method uses ffmpeg to decode the raw video stream (RTSP for Mambo, RTP for Bebop) and convert it into PNG images that OpenCV can process.
Key Characteristics:
ffmpeg must write images to disk.images inside the pyparrot package. You must manually clean this folder after each flight.ffmpeg software and the opencv (cv2) package installed on your system.VisionServer.py to visualize the images.# Requirements: ffmpeg software and opencv packagepyparrot.networking.bleConnection module to manage connections to drones via BLE. This module contains the necessary classes and methods to initialize a BLE session and maintain communication with the drone's BLE interface.To use the pyparrot library with a Bebop, you must first connect your controlling device (laptop, computer, etc.) to the drone's Wi-Fi network. Look for a network named Bebop_number (where number varies by drone). Once connected, use the Bebop.connect() method to establish a connection via the library.
from pyparrot.Bebop import Bebop
bebop = Bebop()
success = bebop.connect(10)Pyparrot requires several core packages and connection-specific libraries.
Core Requirements (All Drones):
untangle: Used to parse XML files in the Parrot SDK.Python 3: Pyparrot only works with Python 3.Connection-Specific Requirements:
zeroconf.pybluez and potentially bluepy.opencv and either ffmpeg or VLC (VLC must be version 3.0.1 or greater).# Core requirement
pip install untangle
# WiFi connection requirement
pip install zeroconf
# BLE connection requirements (Linux only)
sudo apt-get install bluetooth
sudo apt-get install bluez
sudo apt-get install python-bluez
sudo apt-get install python-pip libglib2.0-dev
sudo pip install bluepy
sudo apt-get update