MergeSAM: Unsupervised change detection of remote sensing images based on the Segment Anything Model

TL;DR

MergeSAM is an innovative unsupervised change detection method for high-resolution remote sensing images that leverages the Segment Anything Model to effectively capture complex land cover changes.

Contribution

This paper introduces MergeSAM, a novel approach utilizing SAM with MaskMatching and MaskSplitting strategies for improved unsupervised change detection.

Findings

Effectively detects complex land cover changes

Leverages SAM's segmentation for multitemporal mask construction

Addresses object splitting and merging challenges

Abstract

Recently, large foundation models trained on vast datasets have demonstrated exceptional capabilities in feature extraction and general feature representation. The ongoing advancements in deep learning-driven large models have shown great promise in accelerating unsupervised change detection methods, thereby enhancing the practical applicability of change detection technologies. Building on this progress, this paper introduces MergeSAM, an innovative unsupervised change detection method for high-resolution remote sensing imagery, based on the Segment Anything Model (SAM). Two novel strategies, MaskMatching and MaskSplitting, are designed to address real-world complexities such as object splitting, merging, and other intricate changes. The proposed method fully leverages SAM's object segmentation capabilities to construct multitemporal masks that capture complex changes, embedding the spatial structure of land cover into the change detection process.