Skip to content

README.md

Latest commit

 

History

History
196 lines (141 loc) · 4.67 KB

README.md

File metadata and controls

196 lines (141 loc) · 4.67 KB

🔐 VerifyMe.AI

AI-powered CAPTCHA system using face and hand gesture recognition for secure human verification.

📌 Project Summary

VerifyMe.AI replaces traditional CAPTCHA methods with a smart, real-time human verification system. It uses webcam-based face detection and hand gesture recognition to ensure that only real users — not bots — pass through security checks.

This system is ideal for online exams, secure login systems, or kiosk-based public services. The user is asked to perform a random challenge (e.g., show 2 fingers and smile), and the system verifies both face presence and gesture correctness.

🎯 Key Features

  • ✅ Detects and verifies human face using Haar Cascades
  • ✋ Identifies hand gestures (1–5 fingers) using Mediapipe
  • 🎥 Works in real-time using webcam
  • 🔐 Enhanced security against bots and spoofing
  • 💡 Can be extended to mobile, web, or kiosk systems

🧠 Real-World Applications

Use Case Examples
Online exam portals Facial + gesture login for human candidates
Government portals Secure public access (UIDAI, IRCTC, etc.)
Banking Kiosks Face + hand check before e-KYC
Cybercafés Prevent bots from auto-registering

🛠️ Technologies Used

  • Python 3.8
  • OpenCV – for webcam and face detection
  • Mediapipe – for hand landmark tracking
  • Haarcascade XML – pre-trained face detection model
  • NumPy – image/array operations

📁 Folder Structure

verifyme-ai/
├── app.py                        # Main script to run system
├── requirements.txt              # Dependencies list
├── LICENSE                       # MIT License
├── README.md                     # This file
│
├── haarcascades/
│   └── haarcascade_frontalface_default.xml
│
├── hand_module/
│   └── hand_gesture.py           # Hand gesture detection logic
│
├── assets/
│   └── bala.jpg                  # Test image or user sample
│
└── .venv/                        # Virtual environment (optional)

📥 Installation Guide

✅ Step 1: Clone the Repository

git clone https://github.com/yourusername/verifyme-ai.git
cd verifyme-ai

✅ Step 2: Create a Virtual Environment

python -m venv .venv

✅ Step 3: Activate the Environment

  • Windows:

    .venv\Scripts\activate

  • Linux/macOS:

    source .venv/bin/activate

✅ Step 4: Install Dependencies

pip install -r requirements.txt

▶️ How to Run

python app.py
  1. The webcam will start.
  2. The system will detect your face using Haar Cascade.
  3. A random gesture challenge appears (e.g., "Show 3 fingers").
  4. Mediapipe detects your hand and counts fingers.
  5. If both checks pass → ✅ CAPTCHA Passed

🔍 How It Works (Flowchart)

[ Webcam Feed ]
       ↓
[ Detect Face (Haar Cascade) ]
       ↓
[ Detect Hand (Mediapipe) ]
       ↓
[ Generate Random Gesture Challenge ]
       ↓
[ Match Face + Gesture ]
       ↓
[ Verification Success ]

🧪 Sample Challenge Logic (Python)

import random
challenge = random.choice([1, 2, 3, 4, 5])
print(f"Show {challenge} fingers to continue")

💡 Example Code Snippet: Face Detection (OpenCV)

face_cascade = cv2.CascadeClassifier('haarcascades/haarcascade_frontalface_default.xml')
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(gray, 1.3, 5)

🖐️ Example Code: Hand Landmark Detection (Mediapipe)

with mp_hands.Hands() as hands:
    result = hands.process(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))
    if result.multi_hand_landmarks:
        count_fingers(result.multi_hand_landmarks)

📷 Demo Preview

![demo]image

🚀 Future Enhancements

  • GUI using Tkinter or Streamlit
  • Mobile version using MediaPipe on Android
  • OTP fallback for invalid attempts
  • Smile/blink detection (multi-modal verification)
  • Voice command captcha support

👨‍💻 Author

Balashanmugam
MSc Data Science | AI Enthusiast | Python Developer
📧 balashanmugamrathinam@gmail.com
🔗 LinkedIn Profile

📝 License

This project is licensed under the LICENSE.
You are free to use, modify, and distribute this code for personal or commercial use.

“VerifyMe.AI: Making CAPTCHA smarter, not harder.”