Traffic Monitoring and Analysis System

Automated real-time traffic analysis using Computer Vision and Edge Computing.

This repository contains the source code for my Bachelor's thesis. The system handles the entire data pipeline—from raw RTSP streams at the edge to a centralized dashboard for urban planning insights.

Key features

• Real-time Detection & Tracking: Leverages Computer Vision to identify and follow vehicles in video streams.
• Edge Processing: Optimized to run on Raspberry Pi using GStreamer and Python.
• Data Pipeline: Reliable data transfer from Edge devices to a central server via MQTT.
• Interactive Dashboard: Visualize traffic flow, peak hours, and vehicle counts in real-time.

System Architecture

The system is divided into an Edge processing unit (AI Module) and a central management layer.

Visuals

Detection Pipeline:

Analytics Dashboard:

Tech Stack

• Edge Hardware: Raspberry Pi
• Language: Python
• Vision: GStreamer
• Communication: MQTT
• Database: PostgreSQL
• BackEnd: FastApi
• Visualization: Streamlit