Body Sign Language To Morse Code

• Purpose
• Features
  • Prerequisites
  • Notes
• Project Structure
• How To Run?
• Test
• References
• License

Purpose

This Machine Learning (ML) project converts body sign language gestures into Morse code using Python and image processing techniques. It is designed to process visual input on a client device (laptop) and send results to a server device (Raspberry Pi) to control LEDs that blink Morse code. The division of processing ensures efficient execution, as the Raspberry Pi cannot run certain libraries like mediapipe effectively.

The project provides a creative and practical tool for learning about Python, ML and hardware integration.

Features

• Real-time body gesture recognition using Python and image processing.
• Conversion of recognized gestures into Morse code.
• Transmission of Morse code data from a client to a server via sockets.
• LED control on a Raspberry Pi to display Morse code.
• Modular project structure for ease of testing and future extension.

Prerequisites

• Raspberry Pi 3B or 4B Model
• Python 3.8+
• Camera (for client device)
• Python libraries specified in requirements.txt
• Optional: GPIO setup for controlling LEDs on the Raspberry Pi

Notes

1. Directly test the project by running: python3 test_classifier.py
   Ensure you've completed Step 3 in the How To Run? section before testing.

Project Structure

The project follows this directory structure:

body_sign_language_to_morse_code/
│
├── client/
│   ├── collect_imgs.py
│   ├── create_dataset.py
│   ├── socket_client.py
│   ├── train_classifier.py
│   └── test_classifier.py
│
├── data/
│
├── images/
│   ├── mors_alphabet_for_numbers.jpg
│   ├── real-time-test.jpeg
│   └── the-26-letters-and-10-digits-of-American-Sign-Language-ASL.png
│
├── models/
│   ├── data.pickle
│   └── model.p
│
├── env/
│
├── server/
│   └── socket_server.py
│
├── gitignore
├── README.md
└── requirements.txt

• client/: Contains client and ML codes.
  • collect_imgs.py: Collect images for training
  • create_dataset.py: Create datasets from collected images
  • socket_client.py: Handles communication with the server
  • train_classifier.py: Train the gesture classifier
  • test_classifier.py: Test the trained classifier
• data/: Contains datas created by collect_img.py file.
• images/: Contains visual informations as images for the project.
• models/: Stores trained models and data files.
• env/: Virtual environment.
• server/socket_server.py: Receives data and controls the LEDs.
• requirements.txt: Lists project dependencies.

How To Run?

1. Clone the repository:

git clone https://github.com/abdulkerimakten/body_sign_language_to_morse_code.git
cd body_sign_language_to_morse_code

2. Virtual environment setup (Optional but recommended):

• For Windows:

python3 -m venv env
env/Scripts/activate

• For Linux & MacOS:

python3 -m venv env
source env/bin/activate

3. Install dependencies:

• If you are running repository on hardware platforms such as Raspberry Pi or Jetson, you should install RPi.GPIO library sepereatly; because this library is designed specifically for devices equipped with GPIO pins and may not be compatible with other platforms.

pip install -r requirements.txt
# or
pip3 install -r requirements.txt

4. Run the project in order:

python3 collect_imgs.py

python3 create_dataset.py

python3 train_classifier.py

python3 test_classifier.py

Test

Preview from the real-time test:

References

• Sign language detection with Python and Scikit Learn...

License

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