High-resolution Microscopic Image Dataset of Freshwater Plankton in Japanese Lakes and Reservoirs (FREPJ): I. Zooplankton

Plankton are important organisms that structure food webs in aquatic ecosystems and are also effective environmental indicators. However, the identification and enumeration of these organisms for environmental monitoring is challenging in terms of sustainability and accuracy. To overcome these difficulties, we collected plankton images that would be usable for developing an AI-based plankton monitoring system. As a series of plankton image collections, we first made an image dataset for zooplankton. This dataset contains a total of 88,653 images of 214 freshwater zooplankton taxa collected from 87 lakes and reservoirs located in different areas of the Japanese archipelago. To obtain these images, zooplankton samples collected at various locations were first scanned using an intelligent microscope, and high-resolution photographs containing multiple plankton individuals were taken. Then, each plankton individual was cropped and extracted from the photographs as a single image, classified and labeled with multiple taxonomic ranks (phylum, class, order, family, genus, and species), and stored in the dataset. The present dataset will be useful not only as an atlas of freshwater zooplankton in Japan, but also for the construction, training, and evaluation of an automatic plankton identification and enumeration system based on machine learning.