IAS pilot - Observations of alien plant species using a vehicle-mounted camera system in Flanders (Belgium)

Событие
Последняя версия опубликовано Research Institute for Nature and Forest (INBO) янв. 27, 2026 Research Institute for Nature and Forest (INBO)
Дата публикации:
27 января 2026 г.
Лицензия:
CC0 1.0

Скачайте последнюю версию данных этого ресурса в формате Darwin Core Archive (DwC-A) или метаданных ресурса в форматах EML или RTF:

Данные в формате DwC-A Скачать 13 880 Записи в English (2 MB) - Частота обновления: по мере необходимости
Метаданные в формате EML Скачать в English (23 KB)
Метаданные в формате RTF Скачать в English (17 KB)

Описание

IAS pilot - Observations of alien plant species using a vehicle-mounted camera system in Flanders (Belgium) is an occurrence dataset published by the Research Institute for Nature and Forest (INBO). It is part of the Biodiversa IAS pilot project (2023-2027), that implemented the monitoring of invasive alien plants using a vehicle-mounted camera system in various countries across Europe. Specifically, the CamAlien camera system (Dyrmann et al. 2024) was mounted on a car to capture images of the roadside vegetation. Data have been standardized to Darwin Core. Issues with the dataset can be reported at https://github.com/inbo/ias-pilot/issues.Images from the cameras were processed with image recognition algorithms provided through Pl@ntNet (Espitalier et al. 2025). For the current dataset, only pre-selected species of regional importance were retained, and occurrences (images) with insufficient confidence were filtered out.We have released this dataset to the public domain under a Creative Commons Zero waiver. We would appreciate it if you follow the INBO norms for data use (https://www.inbo.be/en/norms-data-use) when using the data. If you have any questions regarding this dataset, don't hesitate to contact us via the contact information provided in the metadata or via opendata@inbo.be.This dataset was published as open data for the IAS pilot project (funded by Biodiversa+) and the OneSTOP project (funded by Horizon Europe), with technical support provided by the Research Institute for Nature and Forest (INBO).

Записи данных

Данные этого находка ресурса были опубликованы в виде Darwin Core Archive (DwC-A), который является стандартным форматом для обмена данными о биоразнообразии в виде набора из одной или нескольких таблиц. Основная таблица данных содержит 13 880 записей.

Также в наличии 1 таблиц с данными расширений. Записи расширений содержат дополнительную информацию об основной записи. Число записей в каждой таблице данных расширения показано ниже.

Occurrence (core)
13880
Multimedia 
13880

Данный экземпляр IPT архивирует данные и таким образом служит хранилищем данных. Данные и метаданные ресурсов доступны для скачивания в разделе Загрузки. В таблице версий перечислены другие версии ресурса, которые были доступны публично, что позволяет отслеживать изменения, внесенные в ресурс с течением времени.

Версии

В таблице ниже указаны только опубликованные версии ресурса, которые доступны для свободного скачивания.

Как оформить ссылку

Исследователи должны дать ссылку на эту работу следующим образом:

D'hondt B, Steen F, Neukermans A, Adriaens T, Dalby L, Gillespie M A, Mellerup K A, Høye T T, Bonnet P, Joly A, Govaert S (2026). IAS pilot - Observations of alien plant species using a vehicle-mounted camera system in Flanders (Belgium). Version 1.1. Research Institute for Nature and Forest (INBO). Occurrence dataset. https://ipt.inbo.be/resource?r=ias-pilot-be&v=1.1

Права

Исследователи должны соблюдать следующие права:

Публикующей организацией и владельцем прав на данную работу является Research Institute for Nature and Forest (INBO). Насколько это возможно по закону, издатель отказался от всех прав на эти данные и посвятил их Public Domain (CC0 1.0). Пользователи могут без ограничений копировать, изменять, распространять и использовать работу, в том числе в коммерческих целях.

Регистрация в GBIF

Этот ресурс был зарегистрирован в GBIF, ему был присвоен следующий UUID: 9c85f295-cc92-473c-969a-bb5d3eccfd0b.  Research Institute for Nature and Forest (INBO) отвечает за публикацию этого ресурса, и зарегистрирован в GBIF как издатель данных при оподдержке Belgian Biodiversity Platform.

Ключевые слова

Occurrence; CamAlien; invasive species; alien species; non-native species; IAS; plants

Контакты

Bram D'hondt
  • Metadata Provider ●
  • Originator ●
  • Point Of Contact
Research Institute for Nature and Forest (INBO)
Frédérique Steen
  • Originator
Research Institute for Nature and Forest (INBO)
Axel Neukermans
  • Originator
Research Institute for Nature and Forest (INBO)
Tim Adriaens
  • Originator
Research Institute for Nature and Forest (INBO)
Lars Dalby
  • Metadata Provider ●
  • Originator
Aarhus University
Mark A.K. Gillespie
  • Metadata Provider ●
  • Originator
Aarhus University
Kavi Askholm Mellerup
  • Originator
Aarhus University
Toke T. Høye
  • Metadata Provider ●
  • Originator
Aarhus University
Pierre Bonnet
  • Originator
CIRAD, AMAP
Alexis Joly
  • Originator
Inria
Sanne Govaert
  • Metadata Provider ●
  • Originator
Research Institute for Nature and Forest (INBO)

Географический охват

Flanders, Belgium

Ограничивающие координаты Юг Запад [50,705, 4,216], Север Восток [51,072, 4,983]

Таксономический охват

6 non-native species of vascular plants

Genus Reynoutria sp.
Species Ailanthus altissima, Impatiens glandulifera, Robinia pseudoacacia, Rosa rugosa

Временной охват

Дата начала / Дата окончания 2024-07-18 / 2024-09-09

Данные проекта

Biodiversa+ monitoring pilots aim to facilitate the harmonisation of biodiversity monitoring at a transnational level, by testing common protocols and new technologies. The Biodiversa+ IAS pilot project (2023-2027) implemented the monitoring of invasive alien plants using a vehicle-mounted camera system in various countries across Europe.

Название Monitoring Invasive Alien Species with image-based methods
Идентификатор IAS pilot
Финансирование Biodiversa+, the European Biodiversity Partnership supporting excellent research on biodiversity with an impact on policy and society. Biodiversa+ was jointly developed by BiodivERsA and the European Commission as part of the EU Biodiversity Strategy 2030.
Описание района исследования Europe
Описание плана выполнения исследований During several field seasons of the project, the CamAlien camera system (AI Lab, Denmark) was mounted on vehicles to capture images of the vegetation along linear transport infrastructure. Typically, this system was mounted on a car for the monitoring of roadside vegetation. The system was also tested on other types of vehicles (e.g., boats, trains). Plant species were then identified from the images using species recognition algorithms (Pl@ntNet). The focus was specifically on invasive alien plants, in the context of large-scale surveys, rapid detection, and support of species management. This pilot study was conducted in 11 countries across Europe.

Исполнители проекта:

Bram D'hondt
  • Author
Axel Neukermans
  • Author
Lars Dalby
  • Principal Investigator
Mark A.K. Gillespie
  • Author
Kavi Askholm Mellerup
  • Author
Toke T. Høye
  • Principal Investigator
Sanne Govaert
  • Publisher

Методы сбора

The CamAlien system essentially is a camera mounted on a moving car. The system continuously takes photos while driving, even at highway speeds. Images are anonymised for GDPR compliance and analysis is powered by the Pl@ntNet API, which returns species confidence scores. The software applies adaptive recording based on speed, and a human observer is able to tag plant sightings on the go, helping to build a better training set for the image recognition models.

Охват исследования The study aimed to detect a set of predetermined invasive plant species in roadside verges on a regional scale during the field season of several consecutive years.
Контроль качества The species recognition algorithm (Pl@ntNet) returned a confidence value for each identification of a plant species in an image. A threshold value above which a confidence score is considered accurate was determined through expert validation and statistical modelling. Expert botanists annotated a subset of images with a range of confidence scores for presence/absence of the classified species. Presence/absence was then modelled as a binary response in a logistic GLM (General Linear Model) against confidence score. The confidence score providing a 50% probability of presence was then selected as threshold value. The images resulting from this filtering process were not all individually reviewed.

Описание этапа методики:

  1. Partners from the project consortium defined the set of target species (invasive alien plant species of interest), and agreed on a generic sampling plan for the coming field season.
  2. The CamAlien camera system was mounted on a vehicle.
  3. The vehicle was driven along the road infrastructure, with the CamAlien system running.
  4. The resulting images were shared with the lead partner (Aarhus University), for storage and first-level edits.
  5. The images were transferred to Pl@ntNet, which returned classifications (identifications) of the target species in each of the images.
  6. The returned data were processed, by filtering on the confidence level threshold, or any other step deemed necessary to increase reliability.

Библиографические ссылки

  1. Dyrmann M, Skovsen SK, Christiansen PH et al. (2024) High-speed camera system for efficient monitoring of invasive plant species along roadways. F1000Research 2024, 13:360. https://doi.org/10.12688/f1000research.141992.2
  2. Espitalier V, Lombardo JC, Goëau H, Botella C, Høye T T, Dyrmann M, Bonnet P, Joly A (2025). Adapting a global plant identification model to detect invasive alien plant species in high-resolution road side images. Ecological Informatics 89. https://doi.org/10.1016/j.ecoinf.2025.103129

Дополнительные метаданные

Альтернативные идентификаторы 9c85f295-cc92-473c-969a-bb5d3eccfd0b
https://ipt.inbo.be/resource?r=ias-pilot-be