Brahmilipi to Kannada Character Recognition System

An AI-powered character recognition system that converts Brahmilipi script images to Kannada Unicode characters using deep learning and computer vision techniques.

• Deep Learning Model: CNN-based architecture with TensorFlow/Keras for accurate character recognition
• Synthetic Data Generation: Advanced pipeline creating training data with noise injection and geometric transformations
• Web Interface: Flask-based application with real-time image upload and prediction
• Multi-Character Support: Recognizes 7 core Kannada vowels and consonants (ಅ, ಆ, ಇ, ಈ, ಉ, ಊ, ಕ)
• Image Processing: Robust preprocessing pipeline with OpenCV for various image formats

• Backend: Python, Flask
• Machine Learning: TensorFlow, Keras, scikit-learn
• Computer Vision: OpenCV
• Frontend: HTML, CSS, JavaScript (jQuery)
• Data Processing: NumPy, JSON

• Python 3.8 or higher
• pip package manager

1. Clone the repository

   git clone <repository-url>
   cd Bralmilipi_to_Kannada_Translator

2. Install dependencies

   pip install -r requirements.txt

3. Train the model (if not already trained)

   python src/train_model.py

1. Start the Flask server

   python src/main.py

2. Access the application

   • Open your browser and navigate to http://127.0.0.1:5000
   • Upload an image containing Brahmilipi characters
   • Get instant Kannada character predictions

from src.train_model import predict_character, generate_synthetic_images

# Generate test image
synthetic_images, labels = generate_synthetic_images(num_images_per_class=1)
test_image = synthetic_images[0]

# Predict character
predicted_char = predict_character(test_image)
print(f"Predicted character: {predicted_char}")

Bralmilipi_to_Kannada_Translator/
├── src/
│   ├── main.py                 # Flask web application
│   ├── train_model.py          # Model training and prediction
│   ├── preprocess.py           # Image preprocessing utilities
│   ├── utils.py                # Helper functions
│   ├── templates/
│   │   ├── index.html          # Main web interface
│   │   └── display_images.html # Image display page
│   └── static/
│       └── uploads/            # Uploaded images directory
├── data/
│   └── mapping.json            # Character mappings
├── character_mappings.json     # Model character mappings
├── kannada_synthetic_character_model.h5  # Trained model
├── requirements.txt            # Python dependencies
└── README.md                   # Project documentation

• Input Layer: 64x64x1 grayscale images
• Convolutional Layers: 2 Conv2D layers with BatchNormalization
• Pooling: MaxPooling2D for feature reduction
• Regularization: Dropout layers (0.25-0.5) to prevent overfitting
• Output: 7-class softmax classification for Kannada characters

• Training Accuracy: ~85% (with synthetic data)
• Validation Accuracy: ~79%
• Test Accuracy: ~75%

Note: Performance can be improved with real character image datasets

• Currently trained on synthetic data - real character images would improve accuracy
• Limited to 7 characters - can be extended to full Kannada alphabet
• Model accuracy needs improvement with better training data

• Expand character set to complete Kannada alphabet
• Implement real character image dataset collection
• Add data augmentation techniques
• Improve model architecture for better accuracy
• Add batch processing capabilities
• Implement character sequence recognition

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

• TensorFlow team for the deep learning framework
• OpenCV community for computer vision tools
• Flask team for the web framework
• Contributors to the Kannada Unicode standard