AgData — EU Partnership on Agriculture of Data

This portfolio is an internal working tool of the AgData partnership. Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union. Neither the European Union nor the granting authority can be held responsible for them.

agricultureofdata.eu

Co-funded by the European Union — EU Partnership Agriculture of Data
AgDataUC Portfolio
HomeClustersKPIsTagsSign in

UC 97 — UAVWEED

Some details are only shown to signed-in consortium members. Sign in to see them.

Key Information
Title
Very high resolution image dataset synthesis for AI modelling of weed plants
Acronym
UAVWEED
Coordinator
Leibniz Institute for Agricultural Engineering and Bioeconomy (ATB Potsdam) (Germany)Research Institute
Duration
1 months
Budget
Not shown at your access level. Hidden
Maturity
Committed
Source
In Proposal
Use Case Summit Pitch
Health, Pests, Diseases & Protection
Sectors
Arable crop
Data Types
Machine sensor dataEarth observation dataFarm Management Information Systems (FMIS) data
SRIA Activities

3.1.2 Data integration and data quality

  • 3.1.2.1 Data acquisition & re-use framework
  • 3.1.2.4 Reference data sets & non-discriminatory data
  • 3.1.2.6 Boost data re-usability through quality control
  • 3.1.2.8 Context-based curation & curated small data

3.1.4 Applications of AI techniques

  • 3.1.4.1 Identify key reference/training data sets

3.2.4 Data-based solutions for addressing environmental challenges

  • 3.2.4.3 Bridge crop/pasture yield gaps
  • 3.2.4.4 Areas for biodiversity & pollinator conservation

3.3 Data-based solutions for policy-making

  • 3.3.9 New satellites, drones & ground sensors for policy
Themes
AI / ML / Decision SupportEarth observation & remote sensingPlant protection & pest/disease
Partners (2)
  • Leibniz Institute for Agricultural Engineering and Bioeconomy (ATB Potsdam)P1
    Coordinator

    GermanyResearch Institute

  • Leibniz Institute for Agricultural Engineering and Bioeconomy (ATB Potsdam)P2

    GermanyResearch Institute

#OrganisationCountryType
P1Leibniz Institute for Agricultural Engineering and Bioeconomy (ATB Potsdam)Coordinator

*Short summary of your UC* UAVWEED will publish an open, very high-resolution UAV image dataset with expert-validated weed plant annotations in winter wheat crops. The dataset builds on field campaigns, image annotation workflows and weed mapping results from weed-AI-seek and BETTER-WEEDS, and is designed to train and benchmark AI models for low-altitude UAV-based weed detection. Introduction Weeds can reduce crop yield, but they also contribute to biodiversity and ecosystem services. Practical weed management therefore needs reliable information on where weed species occur and which species require action. Today, site-specific control is limited by the lack of large, well-documented, species-level annotated image datasets for robust AI model development. Methodologies Low-altitude RGB UAV imaging in winter wheat crops from very low flight heights (1-3 m altitude) Expert-supported bounding box annotations using CVAT annotation software Quality control through field inventories, expert knowledge and georeferenced ground truth weed assessments. Integration of associated evidence from weed-AI-seek: YOLO-based real-time UAV detection and edge- computing workflows Output Open Access Annotated Weed Image Dataset with high-resolution UAV images and species-level annotations. Technical documentation and annotation guidelines covering data generation, formats, metadata and quality assurance. Impact The dataset supports AI-based weed detection for example for spot spraying, selective weeding and biodiversity-friendly crop protection. It contributes to FAIR agricultural data, improved model generalisation, reduced pesticide use, and European data-space activities linked to the Green Deal, Farm-to-Fork, CAP eco- schemes and biodiversity monitoring. What I can offer Curated UAV image data and annotation experience for weed detection in arable crops. Domain expertise in UAV field campaigns, weed mapping, ground truthing and AI model validation. What I need Support in metadata harmonisation, repository integration and long-term visibility. Exchange with partners on common annotation standards, data governance, and new EU projects.

Source: Annual AgData Use Case Summit 2026 booklet.

Test and experimental facilities (TEF research) data

4.2 Data governance, standards and security (R&I activities)

  • 4.2.4.15 Cybersecurity in agricultural value chains
Germany
Research Institute
P2Leibniz Institute for Agricultural Engineering and Bioeconomy (ATB Potsdam)GermanyResearch Institute