From camera trap to GBIF - Project report for the NLBIF project "Van cameraval tot GBIF"

Final report for the project Van cameraval tot GBIF (nlbif2018.2019.017), funded by the NLBIF-call 2018/2019 which was launched on February 4, 2019. The project ran from July 2019 to June 2021 and developed tools and formats to facilitate sharing and archiving camera trap data. Camera trapping has become one of the most important technologies in conservation and ecological research and is now a well-established, non-invasive method for collecting field data on animal abundance, distribution, behaviour, temporal activity and space use. To facilitate researchers in managing camera-trap projects and annotating and storing camera trap images, Wageningen University & Research (WUR) and the Research Institute for Nature and Forest (INBO) develop and maintain the platform Agouti. In the NLBIF-funded project “Van cameraval tot GBIF” (2019-2021), we developed tools and formats to facilitate sharing and archiving this type of data. The initially suggested Camera Trap Metadata Standard (CTMS) turned out to be insufficient and no longer maintained as a data exchange format. We therefore partnered with the Open Science Conservation Fund (creators of TRAPPER) to develop a new data exchange format: Camera Trap Data Package (Camtrap DP). This format is now hosted by Biodiversity Information Standards (TDWG), reaching a stable release and gaining wider support. We updated Agouti so that users can export their data in this standardized format with a single click. Agouti now also serves public images, and allows to hide images containing humans to comply with privacy laws. To demonstrate the Camtrap DP format, we published an open dataset in this format on Zenodo, containing over 1.1 million images and 61.000 annotations. We also created a conversion script from Camtrap DP to Darwin Core, which expresses the animal observations and associated images into an Occurrence Core and Simple Multimedia Extension. This conversion was used to publish the example dataset as a new dataset to GBIF. Both datasets are licensed under a CC0 waiver. The software code, dataset and especially the Camtrap DP data exchange format have the potential to be widely used beyond Agouti. Overall, the project has made a substantial contribution to the standardized exchange of camera trap data, both within the camera trapping and GBIF community, and we thank NLBIF for the financial support.This project was realized with a financial contribution from Stichting NLBIF - Netherlands Biodiversity Information Facility. NLBIF supports Dutch organizations and researchers in putting biodiversity data online via gbif.org in an open and internationally standardized manner