AutoGeoLabel: Automated Label Generation for Geospatial Machine Learning

TL;DR

This paper presents AutoGeoLabel, a platform-independent method for automatically generating high-accuracy labels for geospatial data using rasterized statistical features, facilitating supervised learning in remote sensing applications.

Contribution

It introduces a novel automated label generation pipeline for geospatial data that achieves high accuracy and is adaptable across different satellite modalities and platforms.

Findings

Achieved ~0.9 accuracy in multi-class label generation

Demonstrated platform independence and adaptability

Validated on dense urban areas with multiple land cover classes

Abstract

A key challenge of supervised learning is the availability of human-labeled data. We evaluate a big data processing pipeline to auto-generate labels for remote sensing data. It is based on rasterized statistical features extracted from surveys such as e.g. LiDAR measurements. Using simple combinations of the rasterized statistical layers, it is demonstrated that multiple classes can be generated at accuracies of ~0.9. As proof of concept, we utilize the big geo-data platform IBM PAIRS to dynamically generate such labels in dense urban areas with multiple land cover classes. The general method proposed here is platform independent, and it can be adapted to generate labels for other satellite modalities in order to enable machine learning on overhead imagery for land use classification and object detection.