Sign Language Translator

A real-time sign language translation system using computer vision and machine learning. This project uses hand tracking and gesture recognition to translate American Sign Language (ASL) letters into text.

ASLTranslator.3.mp4

Features

• Real-time hand tracking and detection
• Sign language gesture recognition
• Support for ASL letters (currently A, B, C)
• Live video feed with visual feedback
• Real-time predictions with confidence scores

Requirements

• Python 3.x
• OpenCV (cv2)
• cvzone
• TensorFlow/Keras
• NumPy

Installation

1. Clone this repository:

git clone [this-repository-url]
cd sign_language_translator

2. Install the required dependencies:

pip install opencv-python cvzone tensorflow numpy

3. Make sure you have a webcam connected to your computer.

Usage

1. Run the main application:

python dataCollection.py

2. Position your hand in front of the camera.
3. The system will detect your hand gestures and display the corresponding ASL letter prediction.

How It Works

The application uses the following components:

1. Hand Detection: Uses cvzone's HandTrackingModule to detect and track hands in real-time.
2. Image Processing:
   • Crops and processes the hand region
   • Normalizes the image to a standard size
   • Maintains aspect ratio for consistent recognition
3. Classification: Uses a trained Keras model to classify hand gestures into corresponding ASL letters.
4. Visual Feedback: Displays the recognized letter and bounding box around the detected hand.

Model Training

The system uses a pre-trained model stored in Model/keras_model.h5. The model was trained on ASL letter gestures and can recognize letters A, B, and C.