Form Digitization Using CV2 and AI 🎥🤖

This project leverages OpenCV and the Gemini 1.5 Flash API for real-time form digitization. It provides an interactive interface to digitize forms, detect hand gestures for form type selection, and overlay extracted data on a visually appealing custom background.

Features 🌟

• Real-Time Video Feed: Captures live input from the webcam.
• Hand Gesture Recognition: Uses finger gestures to select the form type (e.g., Student Card, Challan, General Form).
• Gemini 1.5 Flash API Integration: Extracts text from images of forms uploaded by the user.
• Dynamic Data Display: Displays extracted data (like name, CNIC, department, etc.) directly on the application interface.
• Customizable UI: Implements a styled UI with background overlays and region-specific displays.
• Video Recording: Allows recording of sessions for future reference.
• Multi-Mode Support: Toggle between instruction view, live capture, and extracted text view.

---

Prerequisites 📋

Make sure you have the following installed and set up before running the application:

• Python 3.8+
• OpenCV (cv2)
• cvzone (for hand gesture detection)
• Pillow (for image processing)
• Gemini 1.5 Flash API credentials
• A webcam for live video input

---

Installation 🛠️

1. Clone the repository:

   git clone https://github.com/cyberfantics/form_digitilization.git
   cd form_digitilization

2. Install the required dependencies:

   pip install -r requirements.txt
   

3. Add your Gemini 1.5 Flash API key in the extract.py script.

---

How to Use ▶️

Run the application:

python main.py

Follow the On-Screen Instructions 📋

• Press p: Toggle video recording.
• Press i: View the instructions screen.
• Press c: Enter live capture mode.
• Press s: Send the frame to the Gemini 1.5 Flash API for text extraction.
• Press h: Activate hand gesture detection for form type selection.
• Press q: Exit the application.

---

Hand Gesture Rules ✋

• Five fingers open: Select General Form Mode.
• Two fingers open (peace sign): Select Fee Challan Mode.
• Five fingers closed (fist): Select Student Card Mode.

---

Output 📊

• The extracted data (e.g., Name, CNIC, Gender) is displayed directly on the UI.
• You can view the processed data in live video frames and save the session as a video file.

---

File Structure 📂

• main.py: Core application logic.
• extract.py: Contains API integration for text extraction.
• resources/: Contains images for the UI (e.g., background and instructions).
• requirements.txt: List of dependencies.

---

Screenshots 📸

---

Contributing 🤝

Contributions are welcome! Feel free to submit a pull request or open an issue to suggest improvements.

---

License 📜

This project is licensed under the MIT License.

---

Author

Syed Mansoor ul Hassan Bukhari
GitHub | LinkedIn

---

Happy Digitizing! 🎉