RIce-Net dataset

A computer vision-based framework, River Ice-Network (RIce-Net), uses the USGS nationwide network of ground-based cameras whose images are published through the National Imagery Management System (NIMS), which is available to the public through the Hydrologic Imagery Visualization and Information System (HIVIS; https://apps.usgs.gov/hivis). RIce-Net classifies ice-affected images, segments the ice presence over river surface, calculates the fraction of ice coverage, and automatically generates a near real-time ice flag. The RIce-Net was trained using images from the Milwaukee River near Cedarburg station (https://apps.usgs.gov/hivis/camera/WI_Milwaukee_River_near_Cedarburg) collected in 2023 and tested using images collected in 2024. This repository provides the training data set of the classification and segmentation models.