Direct naar content

You are here:

  1. Home
  2. Algorithms
  3. Mutation recognition on agricultural plots

Please note: The algorithm descriptions in English have been automatically translated. Errors may have been introduced in this process. For the original descriptions, go to the Dutch version of the Algorithm Register.

Mutation recognition on agricultural plots

Netherlands Enterprise Agency

The algorithm on recognition of mutations on agricultural land (reference register) distinguishes parcels with high certainty into two groups: unmutated parcels and mutated parcels. This allows RFO to go through the group of mutated plots manually faster.

Last change on 12th of March 2025, at 14:24 (CET) | Publication Standard 1.0
Publication category
Other algorithms
Impact assessment
DPIA
Status
In use

General information

Theme

Economy

Begin date

2024-01

Contact information

Via contactformulier: https://www.rvo.nl/onderwerpen/contact/formulier

Responsible use

Goal and impact

The aim is to get the latest version as high as possible in an efficient way, by splitting the work stock. We achieve this goal by recognising equal plots, keeping the overview updated.

Plots that may have changed are thus viewed as quickly and well as possible by a human employee.

As a result, RFO has its overview updated much faster, so there are fewer errors in the land overview.

Considerations

The Netherlands has a rule to have the entire survey updated every 3 years.

This means that very many people are involved to determine by hand that the vast majority has not changed. This is a very expensive process to carry out entirely by hand, so the government is spending more money here. The result of mutation recognition is good and fast filtering with no negative financial impact on grant applicants.

Human intervention

When a piece of land looks changed, people check it out. They do this with the sharpest aerial photo. This group is 100% assessed by hand. Where humans see no change, the mutation signal is ignored.

Risk management

The risk of the algorithm seeing a mutation is mitigated by checking them all by hand.

Differences in operation of the algorithm per area type are reviewed based on cultural-historical landscape type. On this, minor adjustments per area are made to the model. This is explained in a technical report and in an operational report.

Legal basis

Up-to-date and correct management of agricultural parcels are crucial for implementing the common agricultural policy in terms of direct payments and conditionalities. This is reflected in various national and international laws and regulations.

National laws and regulations

Dutch National Strategic Plan CAP 2023-2027

Implementation regulation CAP 2023

Regulation of the Minister of Agriculture, Nature and Food Quality of 1 November 2022, no WJZ/22031065.

European legislation and regulations

Various regulations surrounding European agricultural policy and its implementation (see section 2.2.6.)

Links to legal bases

  • Nederlands Nationaal Strategisch Plan (NSP) 2023-2027 : https://www.rijksoverheid.nl/documenten/kamerstukken/2022/02/24/verslag-strategisch-gbl-plan-202
  • Verordening (EU) nr. 2021/2115: https://eur-lex.europa.eu/legal-content/NL/TXT/?uri=CELEX:32021R2115
  • Verordening (EU) nr. 2021/2116: https://eur-lex.europa.eu/legal-content/NL/TXT/?uri=CELEX%3A32021R2116
  • Verordening (EU) nr. 2022/126: https://eur-lex.europa.eu/legal-content/NL/TXT/?uri=CELEX:32022R0126
  • Verordening (EU) nr. 2021/1172: https://eur-lex.europa.eu/eli/dec/2021/1172
  • Verordening (EU) nr. 2021/1173: https://eur-lex.europa.eu/legal-content/NL/ALL/?uri=CELEX%3A32021R1173
  • Verordening (EU) nr. 2022/1317: https://eur-lex.europa.eu/legal-content/NL/TXT/?uri=CELEX:32022R1317

Impact assessment

Pre-DPIA

Operations

Data

  • Current aerial photographs of Imagery
  • Current reference plots

Links to data sources

  • Luchtfoto's: https://service.pdok.nl/hwh/luchtfotorgb/wms/v1_0?&request=GetCapabilities&service=wms
  • Referentiepercelen landbouw: https://www.pdok.nl/-/dataset-agrarisch-areaal-nederland-is-referentiepercelen-geworden

Technical design

Step 1. Generate True Orthophoto

In Imaging's stereo aerial photos, each area is visible in two or more aerial photos.

Using this stereo image, we can see depth in the aerial photo. We use this to create an

aerial photo without inversion (perspective): the True Orthophoto.

Step 2. Segmentation agricultural plots, buildings, water, roads, other

We use an AI model that looks at the aerial photo. The model divides the

aerial photograph and the elevation model into contiguous areas with the following classes:

1. Agricultural parcels

2. Buildings

3. Water

4. Roads

5. Other areas.

Other areas include all parts that do not belong to the other classes, such as trees.

The model is trained to recognise this topography based on examples from the

Basic Registration of Large-scale Topography.

External provider

Readar

Similar algorithm descriptions

  • Mutation recognition

    Municipality of Bergen op Zoom

    Based on photo material, mutations in basic data are recognised

    Last change on 15th of July 2024, at 12:42 (CET) | Publication Standard 1.0
    Publication category
    Other algorithms
    Impact assessment
    Field not filled in.
    Status
    In use

  • Data mask

    Municipality of Zandvoort

    Algorithm recognizes and anonymizes (personal) data and confidential data in documents before they are published.

    Last change on 15th of July 2024, at 7:16 (CET) | Publication Standard 1.0
    Publication category
    Other algorithms
    Impact assessment
    DPIA
    Status
    In use

  • Datamask

    Haarlem city council

    Algorithm recognises and anonymises (personal) data and confidential data in documents before they are published.

    Last change on 15th of July 2024, at 7:15 (CET) | Publication Standard 1.0
    Publication category
    Other algorithms
    Impact assessment
    DPIA
    Status
    In use

  • Detect risks in customs declarations on non-veterinary feed, feed materials and additives

    Customs

    This algorithm helps Customs and its enforcement partners select goods for control on a risk-based basis. Among other things, it uses declaration data from companies and assesses whether or not there are risks of non-compliant non-veterinary feed, feed materials and feed additives.

    Last change on 2nd of April 2025, at 12:48 (CET) | Publication Standard 1.0
    Publication category
    Impactful algorithms
    Impact assessment
    Field not filled in.
    Status
    In use

  • Detect risks in customs declarations for compliance with import and export ban of Mercury Regulation

    Customs

    This algorithm helps Customs to select goods for control based on risks. Among other things, it uses declaration data from companies and assesses whether or not there are increased risks of exporting and importing substances listed in the Mercury Regulation in customs declarations.

    Last change on 2nd of April 2025, at 12:33 (CET) | Publication Standard 1.0
    Publication category
    Impactful algorithms
    Impact assessment
    Field not filled in.
    Status
    In use