🔐 FaceLock - Face Recognition System

A Flask-based real-time face recognition web application that allows users to register their faces and identify previously registered users using OpenCV's Haar Cascade face detector and LBPH Face Recognizer.

✨ Features

• 📷 Live webcam feed
• 👤 Register new users with face capture
• 🧠 Train an LBPH face recognition model
• 🔍 Recognize registered users in real time
• 🎨 Modern cyberpunk-inspired web interface
• ⚡ Built with Flask and OpenCV

---

🛠️ Tech Stack

• Python
• Flask
• OpenCV
• OpenCV Contrib (LBPH Face Recognizer)
• NumPy
• HTML/CSS/JavaScript

---

📁 Project Structure

face-recognition/
│── app.py
│── face_model.py
│── requirements.txt
│── .gitignore
│
├── templates/
│   └── index.html
│
├── static/
│   ├── style.css
│   └── script.js
│
├── dataset/          # Generated face images (ignored)
├── trainer.yml       # Trained LBPH model (ignored)
└── labels.npy        # User labels (ignored)

---

🚀 Installation

Clone the repository

git clone https://github.com/navya2006/face-recognition.git
cd face-recognition

Create a virtual environment

python3 -m venv venv

Activate it:

macOS/Linux

source venv/bin/activate

Windows

venv\Scripts\activate

Install dependencies

pip install -r requirements.txt

---

▶️ Run the application

python app.py

Open your browser and visit:

http://127.0.0.1:5000

---

📸 How to Use

Register a Face

1. Enter your Name
2. Enter a unique User ID
3. Click Capture & Train
4. The application captures your face and trains the recognition model.

Recognize a Face

1. Stand in front of the camera.
2. Click Identify Face.
3. The system predicts the registered person's identity.

---

📦 Requirements

• Python 3.10+ (recommended)
• Webcam
• OpenCV Contrib

Install dependencies with:

pip install -r requirements.txt

---

⚠️ Notes

• The dataset/, trainer.yml, and labels.npy files are generated after registering users and are excluded from version control.
• If the webcam does not appear, ensure camera permissions are enabled for your terminal or IDE.
• If you encounter issues with Haar Cascade files, use a stable version of opencv-contrib-python (e.g. 4.10.0.84).

---