Beat The Culprit

A Python CLI application using OpenCV and Azure Face API to detect and identify criminals from video streams in real-time.

Python OpenCV Azure Face API Twilio SMS CLI

View Source

Overview

Beat The Culprit is a command-line application that combines computer vision with cloud-based facial recognition to identify potential criminals from live video feeds. When a match is found, authorities are immediately notified via SMS, enabling rapid response to security threats.

What makes this project special? The entire application was built in just 200 lines of code, demonstrating that powerful security solutions don’t always require complex implementations.

How It Works

The application follows a streamlined pipeline:

Video Capture: OpenCV captures frames from a video stream (webcam or video file)
Face Detection: Each frame is analyzed to detect human faces using OpenCV’s face detection algorithms
Face Recognition: Detected faces are sent to Azure Face API for identification against a database of known criminals
Alert System: If a match is found, the system triggers an SMS alert to notify responsible authorities via Twilio

Key Features

Real-Time Processing: Continuously monitors video streams for faces
Cloud-Powered Recognition: Leverages Azure’s powerful Face API for accurate identification
Instant Notifications: SMS alerts ensure authorities are notified immediately
Lightweight Design: Minimal dependencies and simple architecture
Command-Line Interface: Easy to deploy and integrate with existing systems

Technical Stack

Component	Technology
Video Processing	OpenCV
Face Detection	OpenCV Haar Cascades
Face Recognition	Azure Face API
SMS Notifications	Twilio API
Runtime	Python 3.x

Architecture

Video Stream → Face Detection → Azure Face API → Match Found? → SMS Alert
                    ↓                                ↓
              No Face Found                    No Match
                    ↓                                ↓
            Continue Monitoring              Continue Monitoring

Use Cases

Security Checkpoints: Monitor entry points for known offenders
Public Venues: Enhance security at events, malls, and transportation hubs
Law Enforcement: Assist police in identifying suspects
Access Control: Prevent unauthorized individuals from entering restricted areas

What I Learned

Building this project was a fantastic learning experience. It taught me:

How to work with real-time video processing using OpenCV
Integration with cloud-based AI services (Azure Cognitive Services)
Building notification systems with Twilio
Writing clean, efficient code that does a lot with minimal complexity

The challenge of keeping the codebase under 200 lines forced me to think carefully about architecture and avoid unnecessary abstractions.