Shape recognition public space

Municipality of Amsterdam

This page provides a general overview of the different types of shape recognition of the Amsterdam municipality within the registry. Shape recognition focuses on recognising specific objects or patterns. Forms of shape recognition within the municipality: -object detection waste; -pressure image; and -image recognition advertising tax.

Last change on 5th of March 2026, at 9:40 (CET) | Publication Standard 1.0

• Publication category: Other algorithms
• Impact assessment: Field not filled in.
• Status: In use

General information

Theme

Traffic

Begin date

2022-07

Contact information

[e-mail]

Responsible use

Goal and impact

Municipalities use various technologies for surveillance and enforcement in public spaces. ANPR (Automatic Number Plate Recognition) is specifically developed to automatically record and recognise license plates. This technology evolved from OCR (Optical Character Recognition). ANPR is used for travel time calculation, traffic enforcement, access control and detection of stolen vehicles.

Unlike ANPR, shape recognition focuses on recognising specific objects or patterns, such as a circle instead of letters or characters. Based on the shape, it identifies what the object is. Shape recognition uses algorithms that analyse visual features and is used, for example, to monitor crowds, signal litter or improve public safety. Think of this as identifying a heavy vehicle as a vehicle class based on its shape.

The distinction is thus in the functionality: ANPR technology is exclusively focused on license plates, while shape recognition is more broadly applicable for visual analysis of the environment. ANPR technology has four applications: (1) environmental zone - control of permitted fuel, (2) zero-emission zone - control of vehicles without fuel engines, (3) heavy traffic zone, and (4) closed through traffic - restricted access zone or traffic filters.

ANPR technology enables more than just shape recognition, as in shape recognition. The extent to which privacy-sensitive information is processed differs from that with shape recognition. For this reason, the municipality of Amsterdam chose to include ANPR technology and shape recognition separately in the national algorithm register.

Considerations

For further information on these algorithms, please consult the register. There you will find separate overviews on object detection waste, crowding image and image recognition advertising tax.

Human intervention

For further information on these algorithms, please consult the register. There you will find separate overviews on object detection waste, crowding image and image recognition advertising tax.

Risk management

For further information on these algorithms, please consult the register. There you will find separate overviews on object detection waste, crowding image and image recognition advertising tax.

Legal basis

For further information on these algorithms, please consult the register. There you will find separate overviews on object detection waste, crowding image and image recognition advertising tax.

Elaboration on impact assessments

For further information on these algorithms, please consult the register. There you will find separate overviews on object detection waste, crowding image and image recognition advertising tax.

Operations

Data

For further information on these algorithms, please consult the register. There you will find separate overviews on object detection waste, crowding image and image recognition advertising tax.

Technical design

The municipality of Amsterdam uses camera monitoring with shape recognition to support public space surveillance and public order enforcement. This technology allows automated detection of objects, such as vehicles and flows of people, without directly identifying individuals. No facial recognition or other techniques leading to direct identification are used. Camera surveillance is carried out under the responsibility of authorised officers, ensuring transparency, proportionality and the least intrusion into personal privacy.

Camera monitoring means the following:

The algorithm sees to the shape recognition of the vehicle category of non-Dutch vehicles. In the context of enforcement on mopeds and mopeds, this algorithm is not active. The detection of violations may involve automated decision-making. At present, however, all violations are always checked by Special Investigation Officers.

In addition, camera monitoring includes remote monitoring and operation of bridges and locks. Here, camera images are necessary to assess whether the situation is safe.

Camera monitoring is also used to monitor and monitor safety in road tunnels and associated tunnel sections, for the purpose of incident detection and promoting safety.

Finally, the technology is used to monitor and - if necessary - influence or adjust traffic flows in the context of:

- ensuring good accessibility of public functions and a smooth flow of traffic in a larger area (accessibility and flow);

- achieving a high-quality public space in which road users feel welcome, safe and comfortable (quality and comfort);

- preventing and managing unsafe situations (safety); and

- contributing to the provision of good and timely information to road users.