TrafficMind AI — Smart Traffic Analytics & Optimization

TrafficMind AI is an AI-based traffic analysis and adaptive signal optimization system developed as an academic project. It helps users understand traffic conditions at road junctions by analyzing uploaded traffic images and videos.

Users can create traffic junctions, add lanes, and configure a Region of Interest (ROI) for each lane. The ROI defines the area of the image that belongs to a particular lane. After uploading an image or video, the system uses a pretrained YOLOv8 model to detect vehicles such as cars, motorcycles, buses, and trucks.

The detected vehicles are analyzed and assigned to their respective lanes using the configured ROI boundaries. The system then calculates the number of vehicles in each lane and generates a weighted traffic score. Different vehicle types have different weights because buses and trucks generally occupy more road space than motorcycles and cars.

Based on the traffic score, each lane is classified into a traffic density level such as Low, Medium, High, or Very High. The system also calculates lane priority by comparing the traffic demand of each lane within the same junction. This helps identify which lanes require more green signal time.

TrafficMind AI uses this information to generate a simulated adaptive traffic signal plan. The plan recommends green, yellow, and red durations for each lane according to the detected traffic demand. The signal optimization is only a software simulation for academic demonstration and does not control real traffic lights, physical devices, CCTV cameras, or road infrastructure.

The project includes a web-based dashboard where users can view total vehicles, lane traffic scores, density levels, priority lanes, recent analyses, junction summaries, and simulated signal cycles. Separate analytics and history pages allow users to review traffic trends and previously processed analyses.

TrafficMind AI also provides a versioned REST API using Django REST Framework. Authenticated users can retrieve junction information, lane data, analysis results, lane-level traffic intelligence, and simulated signal plans. The API also supports triggering the existing image or video processing workflow.

This project is designed for beginners and academic users who want to understand how computer vision, traffic analysis, database systems, and web applications can work together. It uses manually configured lane ROIs and a pretrained YOLOv8 model instead of custom model training.

TrafficMind AI is an upload-based academic demonstration system. It does not support live CCTV streaming, real-time traffic monitoring, physical traffic signal control, IoT hardware, automatic lane detection, vehicle tracking, or unique vehicle counting. Video results are based on sampled frames, and the traffic weights and signal timing values are configurable demonstration parameters rather than certified traffic-engineering values.

Usage:

1. Create a traffic junction.
2. Add and configure lanes.
3. Define lane ROI regions.
4. Upload a traffic image or video.
5. Run vehicle detection.
6. Review lane traffic and congestion results.
7. Generate a simulated signal plan.
8. View results through the dashboard, analytics pages, or API.

Python 3.13 or compatible Python version
Virtual environment
Django 5.2
Django REST Framework
Ultralytics YOLOv8
OpenCV
SQLite
A pretrained YOLOv8 model
Traffic image or video samples
Modern web browser