A high-resolution and efficient system for analyzing and labeling ice images

Authors

  • ABISHEK KANTHAN Chantilly High School
  • PATRICK O’BRIEN James Madison High School
  • Theodore Spanbauer Center for Spatiotemporal Thinking, Computing, and Applications, Department of Geography and Geoinformation Science, George Mason University, Fairfax, VA
  • Chaowei Yang Center for Spatiotemporal Thinking, Computing, and Applications, Department of Geography and Geoinformation Science, George Mason University, Fairfax, VA

DOI:

https://doi.org/10.13021/jssr2023.4001

Abstract

A well-known issue with our environment is global warming and climate change. The Arctic region is a key indicator that can be used to measure the effects of global warming. Arctic CyberInfrastructure (ArcCI) allows researchers and scientists to study this pressing issue through autonomous classification of aerial ice image data from the Arctic regions. In addition, ArcCI aims to be user-friendly and intuitive, featuring a simple interface that allows for quick labeling of ice structures using labels such as ice, shadows, and water. ArcCI was made to improve image management and processing, specifically for ice data. ArcCI primarily uses object-based image analysis (OBIA) methods with integrated machine learning and high spatial resolution data to accurately label images. A key component of the ArcCI infrastructure is its ability to classify images autonomously due to the various data training models used and the export script which allows for the ability to easily share labeled images. ArcCI helps enhance the accuracy and consistency of ice image analysis. This makes it a valuable asset in a diverse range of applications, from climate studies to ice navigation. 

Published

2023-10-27

Issue

Section

College of Science: Department of Geography and Geoinformation Science

Categories