This document provides the standard operating procedure (SOP) for the development of training datasets for urchin detection in NOAA’s Video and Image Analytics for Marine Environments (VIAME), an image-recognition AI tool. It includes instructions on how to manually annotate, record species, and size urchins in benthic imagery and automate the calculation of abundance metrics. This SOP workflow includes 1) using ArcGIS Pro to subsample large orthomosaic images and import subsampled areas into VIAME, 2) manually annotating urchins in benthic imagery using the VIAME tool, and 3) confirming urchin species using underlying imagery via Viscore as outlined in the Viscore SOP.
Alagata et al. (Wed,) studied this question.