rymot-hq/amharic-braille-opencv

# Amharic Braille Detection using OpenCV A computer vision-based solution for automatically detecting and interpreting Amharic Braille characters using OpenCV and deep learning. ## Overview Amharic Braille is essential for visually impaired individuals who read and write in Amharic. This project provides an automated system to: - Preprocess images of Amharic Braille documents - Segment individual Braille characters - Classify and interpret Braille characters using machine learning ## Features - **Image Preprocessing**: Noise reduction, perspective correction, and adaptive thresholding - **Character Segmentation**: Automatic detection and extraction of individual Braille characters - **Classification**: Deep learning model for recognizing Amharic Braille characters - **End-to-End Pipeline**: Complete workflow from image input to text output ## Installation ```bash # Clone the repository git clone github.com cd amharic-braille-opencv # Install dependencies pip install -r requirements.txt ``` ## Usage ### Basic Usage ```python from src.main import BrailleDetector # Initialize detector detector = BrailleDetector() # Process an image result = detector.process_image('path/to/braille_image.jpg') print(result) ``` ### Command Line Interface ```bash # Process a single image python main.py --image path/to/image.jpg # Process multiple images python main.py --batch --input-dir raw_dataset/ --output-dir results/ ``` ## Project Structure ``` amharic-braille-opencv/ ├── src/ │ ├── braille_preprocessing.py # Image preprocessing functions │ ├── braille_segmentation.py # Character segmentation │ ├── braille_classifier.py # ML classification model │ ├── utils.py # Utility functions │ └── main.py # Main pipeline ├── models/ # Trained model files ├── raw_dataset/ # Input images ├── output/ # Processed result …