Parking control

Municipality of Oss

Using a (license plate) image recognition algorithm, check whether parked cars have the right to be parked where paid or permit parking applies.

Last change on 8th of September 2025, at 9:53 (CET) | Publication Standard 1.0

• Publication category: Impactful algorithms
• Impact assessment: DPIA
• Status: In use

General information

Theme

Traffic

Begin date

8-23

Contact information

[e-mail]

Link to publication website

Scanauto - Gemeente Oss

Responsible use

Goal and impact

Check parkers quickly and reliably for compliance with parking regulations (policy).

Considerations

The (license plate) image recognition algorithm makes verification fast and reliable. The algorithm collects license plates and creates anonymised images of the parking location. Surrounding images are for objection/appeal and assessment after-charge from desk (desktop enforcement). Data collection is proportionate to purpose. Collected license plates and images are not used for other purposes.

Human intervention

An enforcer checks on the spot if there is no after-payment and assesses the situation. If there is no special situation, an additional parking tax (fine) follows.

Risk management

Risk is the misrecognition of a license plate. This can happen if a character on the license plate is wrongly recognised. The human post-check extracts reading errors, i.e. wrong suspect reports, thus preventing unjustified additional charges. (fines)

Impact assessment

Data Protection Impact Assessment (DPIA)

Operations

Data

The scan car collects location and license plate data for control of parking rights. License plates and photos of the immediate surroundings of parked cars are recorded for interpretation and processing of objections and fines.

If the license plate has been paid for parking, scan data are deleted immediately. If there is no parking right then data is stored for further processing and sending the fine. Data is also stored for handling any objection. The data is not used for anything else and deleted after the objection period has expired. The applied algorithm detects non-relevant license plates and persons in the photos and makes them unrecognisable.

Technical design

Functional explanation of how the algorithm works and the process:

The scan car travels on public roads in areas where parking is allowed after paying parking tax or with a parking permit.

Different types of cameras on the scan car (infrared, colour and panoramic cameras) scan the license plates of all cars in view of the camera.

A GEO position is linked to each license plate via multiple GPS receivers.

The scanned data is securely transmitted to the management system back office.

Using a geographical map provided by the municipality, the algorithm checks the position of the license plate is in a particular parking area.

The license plate and location data are filtered by the system and checked against the national parking rights database (National Parking Register) queried

If the license plate does NOT appear in the register at the correct location/date/time, a suspect notification is generated in the software and a parking controller goes to the parked car. Assesses the situation and confirms to the enforcer that a fine (additional parking tax assessment) will be sent to the owner of the car.

External provider

Brickyard BV