abdennour1728/DocInfoExtract

Modular document processing pipeline for Algerian birth certificates. Evaluates EasyOCR, Tesseract, PaddleOCR, and Gemini. # DocInfoExtract – Algerian Birth Certificate Reader A modular, open‑source pipeline that extracts structured data from scanned Algerian birth certificates. It detects the document barcode, locates nine standard fields, and reads them using a pluggable OCR engine. ## Features - **Preprocessing** – perspective correction, deskew, normalization - **Barcode‑based layout** – uses the barcode as a spatial anchor for field positions - **Nine extracted fields** - ID Number, Birth Date, Latin & Arabic Names, Birth Place (Baladia / Wilaya), Gender, Father & Mother Names - **Four OCR backends** - **EasyOCR** – free, offline, good for mixed scripts - **Tesseract** – free, offline, requires language data - **PaddleOCR** – free, offline, strong Arabic support - **Google Gemini** – cloud, most accurate, requires an API key - **Three user interfaces** - **Command‑line interface** (CLI) with file picker - **Streamlit web app** with copy‑to‑clipboard and Excel download - **Excel add‑in** – click a button inside Excel to process an image - **Easy configuration** – switch OCR engines via the launcher menu; all settings stored in `.env` --- ## Quick start (Windows) ### 1. Prerequisites - **Python 3.11** - Download from python.org - Check **“Add Python to PATH”** during installation - **Git** - Download from git-scm.com - **Tesseract OCR** (only needed if you plan to use the Tesseract engine) - Install from UB Mannheim’s build - Select **Arabic language data (`ara`)** during installation - Verify with `tesseract --version` ### 2. Setup 1. **Clone the repository** (or extract the CD archive) ```bat git clone github.com cd DocInfoExtract ``` 2. **Run `setup_venv.bat`** – this creates a Python virtual environment and installs all dependencies. 3. **Run `activate_venv.bat`** – this activates the environment and sets up the default OCR engine. 4. **Start the launcher** by typing `run` and hitting Enter. You’ll see a menu: ``` 1. Command Line …