IDcard OCR Contract Filler

A Streamlit application that extracts personal information from Vietnamese ID cards (CCCD/CMND) using OCR (Gemini API or EasyOCR), then automatically fills a telecom contract template (MobiFone) or generates a customizable .docx contract. Supports live preview on the web and export to PDF/PNG/DOCX.

---

🚀 Features

• 📷 Upload an ID card image (.jpg, .jpeg, .png).
• 🔎 OCR extraction using:
  • Google Gemini Vision API (cloud, high accuracy).
  • EasyOCR (local, offline).
• ✏️ Auto-detect fields: Full name, ID number, Date of birth, Gender, Nationality, Place of origin, Residence, Issue date, Expiry date.
• 📝 Editable form: review and adjust extracted fields directly in the web UI.
• 📄 Contract generation:
  • Official MobiFone telecom contract template (with correct formatting and checkboxes).
  • Generic .docx contract for other use cases.
• 👀 Live preview of contract on the web before downloading.
• ⬇️ Export to PNG / PDF / DOCX.

---

📦 Installation

Clone the repository:

git clone https://github.com/SPARK-UNI/IDcard-ocr-to-contract.git
cd IDcard-ocr-to-contract

Install dependencies:

pip install -r requirements.txt

Dependencies include:

• streamlit
• pillow
• numpy
• python-docx
• easyocr
• google-generativeai (for Gemini API)
• opencv-python

---

⚙️ Configuration

1. Gemini API Key

Create an API key in Google AI Studio.

Set the environment variable:

Windows (PowerShell):

setx GOOGLE_API_KEY "AIza..."

Linux / Mac:

export GOOGLE_API_KEY=AIza...

2. Fonts

Make sure you have a Unicode font that supports Vietnamese.
Recommended: Noto Serif.
Place the .ttf file in the assets/ folder, or use system fonts (Arial, Times New Roman).

---

▶️ Usage

Run the app with Streamlit:

streamlit run app.py

Steps:

1. Upload an ID card image.
2. Wait for OCR (Gemini or EasyOCR).
3. Review & edit extracted fields.
4. Preview the filled contract (MobiFone form).
5. Download as DOCX / PDF / PNG.

---

📸 Demo

ID Card OCR → Field Extraction

Auto-filled MobiFone Contract

---

📂 Project Structure

IDcard-ocr-contract-filler/
│── app.py                        # Main Streamlit app
│── services/
│   ├── ocr_service.py            # OCR with Gemini/EasyOCR
│   ├── field_extraction.py       # Regex & normalization
│   ├── contract_form_mobifone.py # Template filling & rendering
│   └── contract_docx.py          # Generic DOCX generation
│── assets/
│   ├── mobifone_template.png     # MobiFone contract form
│   ├── Demo_UI.png               # UI 
│   └── NotoSerif-Regular.ttf     # Font for Vietnamese text (You should add by yourself)
│── requirements.txt
│── README.md

---

📜 License

This project is licensed under the MIT License. You can freely use and modify it.

---

✨ Acknowledgements

• EasyOCR
• Google Gemini API
• Streamlit