A SPATIAL-FREQUENCY MULTI-BRANCH FRAMEWORK FOR BUILDING EXTRACTION FROM OPTICAL AND SAR IMAGES

Multimodal fusion processing of optical and Synthetic Aperture Radar (SAR) remote sensing images is a practical approach to enhance building extraction performance. However, overcoming domain gaps and efficiently integrating features from different sensors remains challenging. In this study, we propose a novel optical-SAR multi-branch multimodal fusion processing framework that achieves efficient building extraction by leveraging both spatial and frequency domains. We utilize an attention mechanism to fuse global and local features of optical and SAR images in the spatial domain. Additionally, we employ wavelet transform to fully integrate multi-scale, multimodal frequency domain features. Extensive experiments demonstrate that the proposed method achieves excellent results on the SpaceNet6 dataset.