pyparrot-control-using-hand-gesture

📖 Description

This project is a proof of concept for controlling a drone using hand gestures. The project is based on the pyparrot library.

The project is divided into two parts:

1. hand_gesture_recognition which is responsible for detecting hand gestures and sending the appropriate command to the drone.
2. drone_control which is responsible for receiving the command from the hand_gesture_recognition and controlling the drone.

This project uses the mediapipe library for hand gesture recognition.

🛠️ Requirements

• Ubuntu
• miniconda

💻 Installation

1. create conda virtual environment by running this command:

$~ make virenv

2. activate the environment by running this command:

$~ conda activate parrot-env

3. install requirements by running this command:

$~ make install

📑 Usage

1- You have to connect to the drone WiFi network. Example: RS_XXXXXXXXXX

2- Run the main.py file in the src directory, using this command: $~ python src/main.py

**BECAREFUL: BY RUNNING THIS COMMAND THE DRONE WILL TAKEOFF **

if you want to test the FingersCounter class, you can run test.py file in the src directory, using this command: $~ python src/test.py

This is the list of commands to control the drone:

Finger	Command
0	Land
1	Hover in place
2	pitch forward
3	pitch backward
4	roll right
5	roll left
6	yaw right
7	yaw left
8	go up
9	go down

📹 Examples

Here are some examples of the hand gesture recognition:

📹 1

📹 2

📹 3

🌀 Configurable Parameters

You can change the following parameters in the config.yaml file:

Parameter	Description
min detection confidence	minimum hand detection confidence
min tracking confidence	minimum hand tracking confidence
take off altitude	altitude to take off to
duration	duration of the command (the bigger the slower)

🧪 Testing

You can test the project by running this command:

$~ make test

📚 FAQ

What type of drones are supported by this project?

This project is based on the pyparrot library. It supports only the Bebop 2 drone.

Is there any specific hardware requirements for using this project?

No, you can use any laptop or desktop computer.

How do I calibrate the hand gesture recognition system?

You can calibrate the hand gesture recognition system using the configurablesparameters in the config.yaml file. Check the Configurable Parameters section for more details.

How do I calibrate the drone control system?

You can calibrate the drone control system using the configurable parameter duration in the config.yaml file. Check the Configurable Parameters section for more details.

What safety precautions should I take when using this project?

You should be careful when using this project. You should be aware of the surroundings and make sure that there are no obstacles in the drone's path. You should also make sure that the drone is not flying too high out of the WiFi range.

Also you should take care when running the main.py file. By running this file, the drone will takeoff. So make sure that you are in a safe place before running this file.

How to test the project without a drone?

For testing the project without a Parrot Bepop 2 drone, you will be testing the hand_gesture_recognition part only. To do that, you can run this command:

$~ python src/test.py

📝 License

This project is licensed under the MIT License - see the LICENSE file for details

📚 References

• pyparrot
• mediapipe

🔼 Contribution

If you want to contribute to this project, you can do it by forking the project and creating a pull request.