AI-Based Vehicle Detection, Tracking and Directional Traffic Counting
A planning-oriented computer-vision workflow for extracting vehicle movement evidence from roadside video near Kawla, Dhaka.
Processed result
Watch on YouTube ↗
InputRoadside video
MethodYOLOv8 + DeepSORT
OutputIN/OUT traffic records
- Detects cars, motorcycles, buses, and trucks.
- Tracks vehicles with persistent IDs.
- Counts directional movement and stores timestamped events.
- Supports traffic-volume and corridor-flow review.
Python · OpenCV · video analytics · urban mobility
Open public-safe Python source ↗







