Crypto Eye - Version 5

A Python-based facial recognition security application designed to protect your workspace privacy in real time. If an unauthorized person is detected viewing the screen while the admin is working, the app instantly locks the screen and displays a security alert message.

Repository: Akilesh-GA/CryptoEye

Overview

Crypto Eye is an intelligent security solution that continuously monitors your workspace environment using real-time facial recognition technology. It identifies authorized users and detects unauthorized access attempts, triggering immediate protective measures to safeguard sensitive information.

Problem Statement

In today's digital workspace, data breaches often occur through unauthorized viewing of sensitive information on screens. Traditional security measures focus on preventing digital access but ignore the physical threat of someone observing your screen.

Solution

Crypto Eye shifts from reactive to proactive security, automatically detecting and preventing unauthorized screen viewing in real-time.

Key Achievements

Real-time detection of unauthorized access attempts
Automatic prevention of data breaches through instant screen locking
Comprehensive audit trail of all security events
Zero-knowledge approach to privacy (facial data processed locally)
Minimal system overhead - runs efficiently in the background

🚀 Features

Core Security Features

1. Real-time Facial Recognition

Continuous monitoring using your webcam
Processes frames at 30 FPS for smooth operation
Utilizes advanced deep learning models for accurate face detection
Handles multiple faces simultaneously

2. Instant Screen Lock

Automatic screen locking when unauthorized access is detected
Prevents sensitive data exposure
Works across Windows, macOS, and Linux
Configurable lock duration and security level

3. Security Alert Notifications

Real-time pop-up alerts notifying the admin
Sound notifications for critical events
Desktop notification integration
Email alerts for critical incidents (configurable)

4. Comprehensive Alert Logging

Maintains detailed logs of all unauthorized access attempts
Timestamp recording for audit trails
Face recognition confidence scores
CSV export functionality for analysis

5. Proactive Security Approach

Shifts from reactive to active problem-solving
Reduces risk of accidental or intentional data exposure
Customizable security policies
Role-based access control

Advanced Features

6. User Management

Admin profile setup and configuration
Multiple authorized user support
User registration with facial enrollment
Session management and auto-logout

7. Configuration Dashboard

Easy-to-use PyQt5 GUI interface
Real-time configuration updates
Security policy customization
System status monitoring

8. Logging & Analytics

Detailed event logging with timestamps
Unauthorized access attempt tracking
System performance metrics
Report generation

9. Database Integration

Local SQLite database for user management
Encrypted storage of facial encodings
Efficient data retrieval and querying

10. Performance Optimization

Lightweight processing
GPU acceleration support
Minimal CPU/memory footprint
Background thread processing

🛠 Tech Stack

Core Technologies

Languages & Frameworks

Python 3.8+ - Core programming language
Flask - Web server and API backend (optional REST endpoints)
PyQt5 - Desktop GUI framework for user interface
Threading - Concurrent processing for real-time monitoring

Computer Vision & AI Libraries

OpenCV (cv2) - Real-time video processing and face detection
face-recognition - Facial recognition and encoding
dlib - Deep learning-based face detection and recognition
NumPy - Numerical computing and array operations
Pillow (PIL) - Image processing and manipulation

Database & Storage

SQLite3 - Lightweight database for user profiles
pickle - Serialization for facial encodings
JSON - Configuration file storage

Additional Libraries

os - System operations and file management
sys - System-specific parameters and functions
time - Time-related functionality
datetime - Date and time handling
subprocess - System command execution for screen locking
logging - Application logging and debugging
csv - CSV export for reports
pyaudio (optional) - Audio notifications
playsound (optional) - Sound alert playback

System Requirements

Hardware

Webcam (minimum HD quality recommended)
Processor: Intel i5/AMD Ryzen 5 or better
RAM: 4GB minimum, 8GB recommended
Storage: 500MB for application and models

Operating Systems

Windows 10/11
macOS 10.14+
Linux (Ubuntu 18.04+, Debian 10+)

Python Version

Python 3.8 or higher

Architecture

Application Architecture

┌────────────────────────────────────────────────────────────┐
│                    USER INTERFACE LAYER                    │
│                    (PyQt5 GUI Components)                  │
│  ┌──────────────────────────────────────────────────────┐  │
│  │  Dashboard | Settings | Alert Logs | User Management│  │
│  └──────────────────────────────────────────────────────┘  │
└────────────────────┬─────────────────────────────────────┘
                     │
┌────────────────────▼─────────────────────────────────────┐
│              BUSINESS LOGIC LAYER                         │
│                                                            │
│  ┌──────────────────────────────────────────────────────┐ │
│  │  Security Manager | Face Recognition Engine         │ │
│  │  Alert Handler    | Session Manager                 │ │
│  │  Configuration Manager                              │ │
│  └──────────────────────────────────────────────────────┘ │
└────────────────────┬─────────────────────────────────────┘
                     │
┌────────────────────▼─────────────────────────────────────┐
│              MONITORING & DETECTION LAYER                │
│                                                            │
│  ┌──────────────────────────────────────────────────────┐ │
│  │  Webcam Capture (Threading)                         │ │
│  │  Face Detection (OpenCV + dlib)                     │ │
│  │  Facial Recognition (face-recognition library)     │ │
│  │  Real-time Analysis & Decision Making              │ │
│  └──────────────────────────────────────────────────────┘ │
└────────────────────┬─────────────────────────────────────┘
                     │
┌────────────────────▼─────────────────────────────────────┐
│              DATA PERSISTENCE LAYER                       │
│                                                            │
│  ┌──────────────────────────────────────────────────────┐ │
│  │  SQLite Database (User Profiles)                    │ │
│  │  Facial Encodings (pickle files / Database)        │ │
│  │  Alert Logs (CSV / Database)                       │ │
│  │  Configuration Files (JSON)                        │ │
│  └──────────────────────────────────────────────────────┘ │
└──────────────────────────────────────────────────────────┘

Workflow Screenshots

Alert Notification

Security alert displayed when unauthorized access is detected:

images/alert.png

The system immediately displays a prominent alert window with:

Alert Title: "UNAUTHORIZED ACCESS DETECTED"
Alert Message: Details about the detected person
Timestamp: When the unauthorized access occurred
Confidence Level: Face recognition confidence score
Action Buttons: Acknowledge, Lock Now, Settings

Alert Log Overview

Comprehensive history of all unauthorized access attempts:

images/alertlog.png

Features include:

Date & Time: Precise timestamp of each alert
Confidence Score: How similar the detected face is to known users
Duration: How long the unauthorized person was detected
Thumbnail: Small preview image of the detected face
Status: Alert severity and resolution status
Export: Button to export logs to CSV

Unauthorized Person Detection - Example 1

Real-world example of unauthorized person being detected:

images/unauthorized_person1.jpg

Shows:

Real-time webcam feed with face detection box
Confidence percentage displayed
"UNKNOWN" label indicating unauthorized access
System immediately triggering lock protocol

Unauthorized Person Detection - Example 2

Another real-world example demonstrating system effectiveness:

images/unauthorized_person2.jpg

Shows:

Multiple detection frames
Face detection accuracy
Real-time alert triggering
Timestamp of detection

Real-time Processing Flow

Application Start
    │
    ▼
Load Configuration & User Profiles
    │
    ├─ Load admin facial encoding
    ├─ Load authorized users list
    ├─ Initialize security policies
    │
    ▼
Start Monitoring Thread
    │
    ├─ Initialize Webcam
    ├─ Start video capture loop
    │
    ▼
Frame Processing (Continuous Loop)
    │
    ├─ Capture frame from webcam
    │
    ├─ Detect faces in frame (OpenCV)
    │         ├─ No face detected → Continue monitoring
    │         └─ Face(s) detected → Continue to next step
    │
    ├─ Generate facial encodings (face-recognition)
    │
    ├─ Compare with known encodings
    │         ├─ Known user detected (confidence > threshold)
    │         │   └─ Update last seen timestamp
    │         │
    │         └─ Unknown user detected
    │             └─ Trigger Security Protocol
    │
    ▼
Security Protocol (When Unauthorized Access Detected)
    │
    ├─ Lock screen immediately
    │
    ├─ Log unauthorized access attempt
    │         ├─ Timestamp
    │         ├─ Face image
    │         ├─ Confidence score
    │         └─ Duration detected
    │
    ├─ Display alert message
    │
    ├─ Play sound notification
    │
    ├─ Send email alert (if configured)
    │
    └─ Update statistics & continue monitoring

User Actions
    │
    ├─ View Logs → Display alert history
    ├─ Configure Settings → Update policies
    ├─ Register User → Add facial encoding
    ├─ Emergency Stop → Disable monitoring
    └─ View Dashboard → System status

Installation

Prerequisites

System Requirements

Python 3.8 or higher installed
pip (Python package manager)
Webcam or external camera connected
Administrator/sudo access for screen locking functionality

Check Python Version

python --version
# or
python3 --version

Step-by-Step Installation

Step 1: Clone the Repository

# Using HTTPS
git clone https://github.com/yourusername/crypto-eye.git
cd crypto-eye

# Or using SSH
git clone git@github.com:yourusername/crypto-eye.git
cd crypto-eye

Step 2: Create Virtual Environment

# Windows
python -m venv venv
venv\Scripts\activate

# Linux / macOS
python3 -m venv venv
source venv/bin/activate

Step 3: Install Dependencies

# Upgrade pip first
pip install --upgrade pip

# Install all required packages
pip install -r requirements.txt

Step 4: Download Face Recognition Models

The first time you run the application, it will automatically download the required pre-trained models.

# Optional: Pre-download models
python -c "import face_recognition; print('Models downloaded successfully')"

Step 5: Initialize Application

# Run setup to create necessary directories and database
python crypto_eye.py --setup

Step 6: Register Admin User

# Launch the application
python crypto_eye.py

The GUI will guide you through:

Setting up the admin account
Capturing facial encoding
Configuring security policies

Platform-Specific Instructions

Windows

Install Visual C++ Build Tools (if not already installed)
- Required for dlib compilation
- Download from: https://visualstudio.microsoft.com/visual-cpp-build-tools/
Install dependencies
```
pip install -r requirements.txt
```
Run application
```
python crypto_eye.py
```

Usage Guide

Basic Usage

1. Launch the Application

python crypto_eye.py

Confidence Scoring

Facial recognition uses confidence levels:

0.95+ (95%): Extremely confident - Authorized
0.85-0.94 (85-94%): Confident - Authorized
0.75-0.84 (75-84%): Moderately confident - May require confirmation
0.50-0.74 (50-74%): Low confidence - Likely unauthorized
<0.50 (<50%): Very low confidence - Treat as unauthorized

Version History

v5.0.0 (Current)

Release Date: August 2024
Features:
- Real-time facial recognition monitoring
- Instant screen lock on unauthorized access
- Comprehensive alert logging and export
- Multi-user support
- Advanced security policies
- Email notifications
- Database-backed user management
- Cross-platform support (Windows, macOS, Linux)

v4.0.0

Added email notifications
Improved facial recognition accuracy
Better error handling

v3.0.0

Initial PyQt5 GUI
Database integration
Alert logging

v2.0.0

Basic facial recognition
Command-line interface

v1.0.0

Initial release
Proof of concept

Acknowledgments

face-recognition library: By Adam Geitgey
OpenCV: OpenCV team and contributors
dlib: By Davis E. King
PyQt5: Riverbank Computing Limited
Community: All contributors and users

requirements.txt

# Core Dependencies
Python==3.8+

# Computer Vision & Face Recognition
opencv-python==4.8.1.78
opencv-contrib-python==4.8.1.78
face-recognition==1.3.5
dlib==19.24.2
numpy==1.24.3

# GUI Framework
PyQt5==5.15.9
PyQt5-sip==12.13.0

# Database
SQLAlchemy==2.0.21
sqlite3  # Built-in

# Web Framework (Optional)
Flask==2.3.3
Flask-RESTful==0.3.10
Flask-CORS==4.0.0

# Image Processing
Pillow==10.0.0
scikit-image==0.21.0

# Data Processing & Utilities
pandas==2.0.3
scipy==1.11.3

# Configuration & Logging
python-dotenv==1.0.0
pyyaml==6.0.1
python-logging-loki==0.3.2  # For centralized logging

# Email & Notifications
python-telegram-bot==20.3  # For Telegram alerts
google-auth-oauthlib==1.0.0  # For Gmail
google-auth-httplib2==0.1.1

# Audio Notifications
playsound==1.2.2
pydub==0.25.1

# System Operations
pyautogui==0.9.53
psutil==5.9.5
pyperclip==1.8.2

# Testing
pytest==7.4.2
pytest-cov==4.1.0
pytest-mock==3.11.1
pytest-qt==4.2.0
mock==5.1.0

# Code Quality
black==23.9.1
flake8==6.1.0
pylint==3.0.2
mypy==1.5.1

# Documentation
Sphinx==7.2.6
sphinx-rtd-theme==1.3.0

# Scheduling & Tasks
APScheduler==3.10.4

# Performance
numba==0.57.1  # JIT compilation for performance-critical code

# Security
cryptography==41.0.4
passlib==1.7.4
python-jose==3.3.0

# Development Tools
pre-commit==3.4.0
black==23.9.1
isort==5.12.0

# Additional Utilities
tqdm==4.66.1  # Progress bars
requests==2.31.0  # HTTP requests
Werkzeug==2.3.7  # WSGI utilities

Name		Name	Last commit message	Last commit date
app		app
embeddings		embeddings
images		images
utils		utils
README.md		README.md
requirements.txt		requirements.txt

Latest commit

History

Folders and files

Repository files navigation