Hand Sign Detection System

This project is designed to detect hand signs using a webcam and convert them to text and audio. It includes the following features:

1. Hand Sign Detection: Using OpenCV and a deep learning model to detect hand signs in real-time.
2. Text-to-Speech (TTS): The detected hand sign is converted into speech using a text-to-speech engine (pyttsx3).
3. Model Training: You can train the model using a dataset of hand signs collected via webcam.
4. Real-time Detection: The system detects and classifies hand signs in real-time.

---

Features

• Hand Detection: Uses OpenCV's hand tracking module to detect and capture hand gestures.
• Model for Classification: A CNN model trained to classify hand signs.
• Text-to-Speech: Detected signs are spoken using pyttsx3 for accessibility.
• Data Collection: You can collect data by pressing the s key to save images for training.

---

Prerequisites

• Python 3.8, 3.9, or 3.10 (TensorFlow 2.13+ supports these versions)
• OpenCV
• TensorFlow
• pyttsx3

---

Setup Instructions

1. Clone the repository

git clone https://github.com/your-repository/hand-sign-detection.git
cd hand-sign-detection

2. Create a Virtual Environment

It's a good practice to use a virtual environment for managing dependencies. Run the following commands to set up and activate a virtual environment:

Windows

python -m venv venv
.�env\Scripts�ctivate

macOS/Linux

python3 -m venv venv
source venv/bin/activate

3. Install Required Packages

After activating the virtual environment, install the required dependencies:

pip install -r requirements.txt

4. Dataset Collection

To collect data for training the model, run the data_collection.py file. This will allow you to capture hand sign images by pressing the s key.

python data_collection.py

The images will be saved in the specified folder (Data/1).

5. Train the Model

After collecting sufficient data, you can train the model by running the following command:

python train_model.py

The model will be saved in the Model directory as keras_model.h5.

6. Test the Model

Once the model is trained, you can test the hand sign detection system by running:

python test_model.py

This will open a webcam window that will detect hand signs and speak the detected signs.

---

File Structure

hand-sign-detection/
│
├── Data/               # Folder containing collected images for training
│
├── Model/              # Folder to save the trained model and labels
│   ├── keras_model.h5  # The trained model
│   └── labels.txt      # File containing the labels
│
├── data_collection.py  # Script to collect hand sign data
├── train_model.py      # Script to train the model
├── test_model.py       # Script to test the hand sign detection
└── README.md           # This file

---

Troubleshooting

• No webcam access: Ensure your webcam is correctly connected and not being used by other applications.

• Missing dependencies: If any libraries fail to install, try upgrading pip and reinstalling the dependencies:

  pip install --upgrade pip
  pip install -r requirements.txt

---

requirements.txt

opencv-python==4.7.0.72
pyttsx3==2.90
tensorflow==2.13.0
numpy==1.23.5
cvzone==1.5.0
Pillow==9.5.0

---

Notes

• Ensure you have a folder Data/1 to store images during the data collection process.
• Adjust the paths in the code if you're using a different directory structure for storing data or models.