In wet natural areas, invasive plants often spread rapidly and go unnoticed. To better protect these hard-to-access habitats and support restoration efforts, geoinformatics researchers at Osnabrück University, in collaboration with Osnabrück University of Applied Sciences and the TERRA.vita Nature and Geopark, are launching the “MapAid” project. Using camera-equipped drones and artificial intelligence (AI), the researchers aim to take nature conservation to new heights.
The so-called “wet tall-grass communities” along watercourses offer valuable biological diversity. At the same time, they are increasingly threatened by invasive neophytes—that is, non-native plant species. The abundant water supply and rivers as natural transport routes create ideal conditions for the rapid spread of these highly competitive species.
Early detection is crucial for containment, but it is nearly impossible to achieve using conventional methods. Due to tall vegetation, fluctuating water levels, and their remote locations, these areas are difficult even for experienced mappers to access and are often only accessible on a sporadic basis. By the time the plants are discovered during a field survey, they are usually already firmly established or have already produced seeds, making their control costly and labor-intensive.
This is where the “MapAid” project comes in. Camera-equipped drones will fly over the study areas using high-resolution multispectral cameras, thereby collecting data over large areas without damaging the sensitive ecosystems through on-the-ground surveys.
To automatically distinguish the immense variety of plant species in the aerial images, the project partners are pursuing an innovative approach to training artificial intelligence: In the greenhouses at Osnabrück University of Applied Sciences, individual neophytes and native perennials are specifically cultivated. Throughout their entire growth phase, the plants are regularly photographed with multispectral cameras. Using these representative laboratory images, the AI learns the specific appearance of the plants so that it can then accurately identify them in the complex drone images taken in the wild—even in the early stages of growth.
“The earlier we detect invasive plant species, the more precisely and effectively we can combat their spread,” says Dr. Thomas Jarmer of Osnabrück University. “By combining drone data with artificial intelligence, we aim to create a foundation that will allow us to monitor large and hard-to-access areas more efficiently,” adds Lucas Wittstruck from the Remote Sensing and Digital Image Processing research group at Osnabrück University.
The “MapAid” project aims to help make the monitoring of invasive species more objective, comprehensive, and resource-efficient. Regular aerial surveys could help identify new occurrences early on and plan targeted control measures in the future. The goal is to contain invasive species as much as possible before they spread over large areas and displace native plant communities.
The project, which begins on September 1, 2026, is funded by the German Federal Environmental Foundation (DBU) with a grant of approximately 173,000 €.
Additional information for the media:
Dr. Thomas Jarmer, Osnabrück University
Institute of Computer Science
thomas.jarmer@uni-osnabrueck.de
