An intelligent PDF cover page analyzer powered by MoondreamAPI and Streamlit
- 📊 Extract cover pages from PDF documents
- 🔍 Detailed visual analysis of cover pages
- 🤖 AI-powered content recognition
- 💫 Real-time processing and results
- 🎨 Clean and intuitive user interface
- 🧪 Comprehensive unit testing
- Clone the repository
- Install dependencies:
pip install -r requirements.txt- Get your Moondream API key
- Run the application:
streamlit run app.py- Streamlit - Web application framework
- MoondreamAPI - AI-powered image analysis
- PyMuPDF - PDF processing
- Pillow - Image handling
├── .streamlit/ # Streamlit configuration files (will be generated once you excecute the program)
├── images/ # Application screenshots and assets
├── prompts/ # AI analysis prompt templates
├── responses/ # JSON files storing analysis results
├── unit_tests/ # Unit tests and test results
│ ├── test_utils.py # Test cases for utility functions
│ ├── run_tests.py # Test runner with JSON reporting
│ └── test_results.json # Detailed test execution results
├── utils.py # Utility functions and helpers
└── app.py # Main application file
- Enter your Moondream API key in the application
- Start analyzing PDF cover pages
- Customize the analysis prompt in
prompts/prompt.md
Run the unit tests:
python unit_tests/run_tests.pyTest results will be saved in unit_tests/test_results.json with detailed execution information.
Feel free to contribute to this project! Open an issue or submit a pull request.
ToDo's:
- Improving prompts to get even more detailed descriptions on PDFs
- Getting a smooth an smooth way solution to analyze the whole content of the PDF while not spending to many api calls, e.g: analyzing the context of the whole book without making 300 pages/ api calls, and spending less calls.
Created with ❤️ by Eduardo Jose Barrios Garcia (@edujbarrios)
This project is licensed under the MIT License - see the LICENSE file for details.