Repository files navigation

An OCR application that extracts text from images.

• Extract text from uploaded images
• Process multiple image formats (PNG, JPG, JPEG, GIF, WEBP)
• User-friendly Streamlit interface
• RESTful API endpoints
• Integration with Langchain for advanced text processing
• Together AI Vision model integration

• Python 3.12 or higher
• Poetry package manager
• Together AI API key

1.Clone the repository:

   git clone https://github.com/yourusername/ImageTextExtractor.gitcd ImageTextExtractor

2.Install dependencies using Poetry:

   poetry install

1.Start the FastAPI backend:

   poetry run python main.py

2.In a new terminal, launch the Streamlit interface:

   poetry run streamlit run ui.py

3.Open your browser and navigate tohttp://localhost:8501

4.Enter your Together AI API key

5.Upload an image and wait for the results

The application exposes a REST API endpoint for OCR processing.

Request:

• URL:http://localhost:8000/ocr
• Method:POST
• Content-Type:multipart/form-data

Parameters:

• file: Image file (supported formats: PNG, JPG, JPEG, GIF, WEBP)
• api_key: Together AI API key
• system_prompt: (Optional) Custom prompt for the vision model

Example using curl:

curl -X POST http://localhost:8000/ocr \-F"file=@/path/to/your/image.jpg" \-F"api_key=your_together_ai_api_key" \-F"system_prompt=Convert the provided image into text"

Response:

poetry run pytest

The application uses the following configurations (defined inconfig.py):

• LOGGING_LEVEL: Default is "INFO"
• SUPPORTED_IMAGE_TYPES: [".png", ".jpg", ".jpeg", ".gif", ".webp"]
• TOGETHER_MODEL_NAME: "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo"

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

• Together AI for providing the vision model
• Langchain for the AI integration framework
• Streamlit for the user interface
• FastAPI for the REST API implementation