Real-time GeoAI for high-resolution mapping and segmentation of arctic permafrost features: the case of ice-wedge polygons
Year: 2022
Type: conference-paper
Abstract: This paper introduces a real-time GeoAI workflow for large-scale image analysis and the segmentation of Arctic permafrost features at a fine-granularity. Very high-resolution (0.5m) commercial imagery is used in this analysis. To achieve real-time prediction, our workflow employs a lightweight, deep learning-based instance segmentation model, SparseInst, which introduces and uses Instance Activation Maps to accurately locate the position of objects within the image scene. Experimental results show that the model can achieve better accuracy of prediction at a much faster inference speed than the popular Mask-RCNN model.
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Funder: National Science Foundation
Awards: MCA: Career Advancement in Polar Cyberinfrastructure: Permafrost Feature Mapping and Change Detection using Geospatial Artificial Intelligence and Remote Sensing, NNA Track 1: Collaborative Research: The Permafrost Discovery Gateway: Navigating the new Arctic tundra through Big Data, artificial intelligence, and cyberinfrastructure, Collaborative Research: CyberTraining: Implementation: Medium: Cyber2A: CyberTraining on AI-driven Analytics for Next Generation Arctic Scientists, GeoAI for Terrain Analysis: A Deep-Learning Approach for Landform Feature Detection
OpenAlex ID: W4308986509
