Abstract
Plankton are a key component in marine food webs and are vulnerable to climate-driven changes. Their phenology, distribution, and body size influence the biological carbon pump and higher trophic interactions. Plankton imaging enables high-frequency observations of community composition and size structure, but operational workflows for processing, validating and publishing large in situ image collections remain limited. This work describes the deployment of the Plankton Imager 10 (Pi-10) aboard the R/V Simon Stevin in the southern North Sea and presents an open pipeline for Pi-10 data processing, classification, validation and standards-based publication. An open-access data processing pipeline is presented, combining geotagged imaging, deep learning classification, and image-based metrics with applicability beyond the Belgian part of the North Sea. Best practices for documenting Pi-10 datasets in Darwin-Core Archives (DwC-A) are presented. Although developed and validated for Pi-10 imagery, the workflow is designed as a modular framework that can be adapted to other plankton imaging sensors.