Geospatial Trail Mapping
Jan 1, 2022
·
1 min read
This project developed a machine-learning pipeline for large-scale geospatial data processing and automated trail detection from noisy GPS traces.
- Designed a trail-network construction algorithm using growing self-organizing maps
- Processed noisy public GPS traces into cleaner map structure
- Implemented efficient data structures to improve computation performance
- Received the IEEE SIEDS Best Paper Award for the resulting publication
