Skip to content
Think Transportation Blog

Computer Vision in Intelligent Transportation Systems

2 min read

Cameras are already everywhere on our roads. Computer vision turns their video into traffic data: it detects and classifies every vehicle and pedestrian, follows their paths and measures how the traffic behaves, without counting by hand. It is one of the building blocks of an Intelligent Transportation System.

From video to traffic data

At Think Transportation we built our own computer vision application, Vision Genius. Its video analytics pipeline locates and categorises vehicles and pedestrians in traffic videos, tracks their movements and extracts parameters such as total and interval counts, lane counts, traffic flow and density (see Automated Extraction of Micro Traffic Parameters).

What it makes possible

Why it matters

Manual traffic counts are slow and costly, and they cover only a few locations for a few hours. Video analytics can cover more locations for longer and repeat the analysis whenever it is needed, so the traffic studies and designs built on the data rest on a fuller picture. Together with our traffic survey app and real-time geospatial data collection app, it brings field data into the office faster.

Want to turn your camera footage into traffic data? Contact our team.

Discuss Your Project with Us

Start Discussion