ICCV 2021 Workshop on LUAI · 2021
A Framework for Semi-automatic Collection of Temporal Satellite Imagery for Analysis of Dynamic Regions
We detail a semi-automatic pipeline for collecting and labeling temporal satellite imagery using OpenStreetMap metadata, reducing manual annotation while surfacing dynamic regions of interest. The workflow feeds downstream change-detection and classification systems, accelerating monitoring initiatives such as construction tracking.
Highlights
- Combined imagery scraping, polygon filtering, and labeling UI into a single reproducible Python pipeline.
- Enabled rapid refresh of construction monitoring datasets without hand-tracing every frame.
Artifacts & reproduction
Semi-automatic satellite data ingestion plus labeling UI for monitoring changing regions.
- Generate OpenStreetMap-guided scrape manifests for temporal imagery.
- Label construction phases with the included lightweight annotation app.
- Export train/val/test splits for change-detection baselines.
- Demonstrates end-to-end dataset design, labeling tooling, and export pipelines for remote sensing change detection.