Land Cover Segmentation with Sparse Annotations from Sentinel-2 Imagery

TL;DR

This paper introduces SPADA, a novel framework for land cover segmentation that effectively utilizes sparse annotations and domain adaptation, significantly improving accuracy over existing methods in environmental mapping tasks.

Contribution

The paper presents SPADA, a new approach that combines sparse annotations with domain adaptation for improved land cover segmentation accuracy.

Findings

SPADA achieves a mean IoU of 42.86 on Urban Atlas.

SPADA attains an F1 score of 67.93 on LUCAS.

The method outperforms existing semantic segmentation approaches.

Abstract

Land cover (LC) segmentation plays a critical role in various applications, including environmental analysis and natural disaster management. However, generating accurate LC maps is a complex and time-consuming task that requires the expertise of multiple annotators and regular updates to account for environmental changes. In this work, we introduce SPADA, a framework for fuel map delineation that addresses the challenges associated with LC segmentation using sparse annotations and domain adaptation techniques for semantic segmentation. Performance evaluations using reliable ground truths, such as LUCAS and Urban Atlas, demonstrate the technique's effectiveness. SPADA outperforms state-of-the-art semantic segmentation approaches as well as third-party products, achieving a mean Intersection over Union (IoU) score of 42.86 and an F1 score of 67.93 on Urban Atlas and LUCAS, respectively.