Research portal

Dutch

An iNaturalist-Pl@ntNet-workflow to identify plant-pollinator interactions – a case study of Isodontia mexicana

Research output: Contribution to journalArticle

  • Nadja Pernat
  • Tom August
  • Quentin Groom
  • Daniyar Memedemin
  • Lien Reyserhove
In this study, we investigated the possibility of inferring information about interactions between plants and pollinators from iNaturalist images. Insects are often photographed on plants, particularly on flowers, where they can be easily observed and stay still. The platform also provides users with an observation field to record the plant species the pollinator is visiting, but what can you do if you are not a botanist and cannot identify the flower? This can be remedied by apps that identify plants using computer-based image recognition, such as Pl@ntNet, hereafter referred to as plantNet or simply the App. By sending an image to the App, the user receives a list of suggestions with probability scores for candidate species based on image recognition by a convolutional neural network. Theoretically, it should be possible - the main idea of this study - to feed iNaturalist photographs of specific pollinators into plantNet to obtain information about which plants they are visiting.
Original languageEnglish
JournalBioHackrXiv
Issue numberMay
Pages (from-to)1-10
Number of pages10
Publication statusPublished - 2022

Projects

Log in to Pure