Traffic Counting Service

The Traffic Counting service is the core measurement capability of AITracker. It enables automatic counting and classification of both vehicles and pedestrians from video recordings. The system tracks every object crossing your defined measurement cross-sections or passing through intersection relationships.

Measurement Types

Cross-Section Counting

Cross-sections were the first service offered by AITracker. The system counts all vehicles and pedestrians crossing a defined line on the road, regardless of:

• Time of day (day/night operation)

• Weather conditions (rain, fog, snow)

• Traffic density

Each cross-section can be configured with:

• Direction of movement (one-way, two-way, bidirectional)

• Multiple lanes

• Custom time intervals for aggregation

Intersection Counting

Since November 2021, AITracker supports automatic counting at roundabouts and intersections. You define the entrances and the relationships (origin-destination pairs) you want to measure.

Supported intersection types:

• Standard 3-way and 4-way intersections

• Multi-entrance intersections (5+ entrances)

• Roundabouts

• Complex junctions

What Can Be Counted?

Vehicles

AITracker recognizes and classifies all standard vehicle types:

• Bicycles and e-bikes

• Motorcycles and mopeds

• Passenger cars (with or without trailers)

• Delivery vans and buses

• Trucks (light, medium, heavy)

• Articulated trucks and road trains

• Buses (urban, coach)

• Special vehicles (tractors, agricultural machinery)

Pedestrians

The system also counts pedestrian traffic, which is particularly useful for:

• Urban planning studies

• Pedestrian crossing analysis

• Public space usage monitoring

• Shopping area flow analysis

Recognized pedestrian categories:

• Pedestrians (walking)

• Cyclists

• Scooter riders

• People with baby strollers

• Special-needs pedestrians

• Traffic-assistive devices

Note: Pedestrian counting presents additional challenges due to frequent trajectory intersections, occlusions, and high similarity between tracked objects. The measurement error for pedestrians is typically higher than for vehicles.

For a complete list of recognized categories and classification schemes, see the Classifications documentation.

Measurement Accuracy

Measurement Type	Accuracy	Processing Speed*
Cross-section (vehicles)	97%	x5
Intersection traffic	90%	x3
Pedestrian traffic	~90%	x3

* Processing speed x5 means 24 hours of video are processed in approximately 5 hours.

Use Cases

• General Traffic Measurements (GPR) - Official traffic surveys according to Polish and international standards

• Origin-Destination studies - Track traffic flows through complex junctions

• Before/After studies - Measure traffic changes due to road modifications

• Peak hour analysis - Identify traffic patterns and congestion points

• Urban planning - Data for traffic modeling and infrastructure planning

• Pedestrian flow analysis - Count foot traffic in urban areas, at crossings, or in public spaces

• Public transport studies - Measure passenger exchange at stops

Generated Reports and Data

Output Files

File	Description
Quantity.csv / Quantity.xlsx	Traffic counts aggregated by time intervals and vehicle categories
History.csv	Complete record of all tracked objects with timestamps
Report.docx	Generated measurement report following GPR format
Diagrams	Daily traffic profiles, category distributions, cartograms

Diagrams and Visualizations

AITracker automatically generates:

• Daily traffic profiles - Hourly distribution charts

• Category distribution - Pie charts showing vehicle class breakdown

• Intersection cartograms - Visual representation of traffic flows at junctions

• Direction arrows - Movement direction visualization on video frames