Developing Cloud-based Image Classification Management and Processing Service for High Spatial Resolution Sea Ice Imagery
Climate change is gradually becoming more of a world issue as global temperatures rise. Rising temperatures have many negative effects on the environment, and the melting of sea ice is a clear indicator for understanding climate change. Measuring the rate at which sea ice melts is challenging for re...
| Main Authors: | , , , , |
|---|---|
| Format: | Article in Journal/Newspaper |
| Language: | unknown |
| Published: |
Mason Publishing
2022
|
| Subjects: | |
| Online Access: | https://journals.gmu.edu/index.php/jssr/article/view/3253 https://doi.org/10.13021/jssr2021.3253 |
| Summary: | Climate change is gradually becoming more of a world issue as global temperatures rise. Rising temperatures have many negative effects on the environment, and the melting of sea ice is a clear indicator for understanding climate change. Measuring the rate at which sea ice melts is challenging for researchers because of the scale at which melting occurs and the presence of various ice thicknesses. Machine learning has been an essential tool that we can use for this purpose. Our lab found that there were no services that provide reliable and efficient on-demand batch image processing for sea ice classification. This led the lab to create the Arctic Cyberinfrastructure (ArcCI) tool to address these issues. The interns have worked on analyzing the algorithms of ArcCI through comparison with other projects and with runtime experiments. We have also worked to prepare the datasets of sea ice images pertaining to the satellite imagery in order to learn about the most efficient way to organize data for training. This work will help improve the algorithms used for image classification and increase the accuracy of the existing model. Moving forward, the lab would like to implement the deep learning model into the platform and further increase classification accuracy under dark lighting contexts. |
|---|