The precise building extraction from high-resolution remote sensing images holds significant application for urban planning, resource management, and environmental conservation. In recent years, deep neural networks (DNNs) have garnered substantial attention for their adeptness in learning and extracting features, becoming integral to building extraction methodologies and yielding noteworthy performance outcomes. Nonetheless, prevailing DNN-based models for building extraction often overlook spatial information during the feature extraction phase. Additionally, many existing models employ a simplistic and direct approach in the feature fusion stage, potentially leading to spurious target detection and the amplification of internal noise. To address these concerns, we present a multi-scale attention network (MSANet) tailored for building extraction from high-resolution remote sensing images. In our approach, we initially extracted multi-scale building feature information, leveraging the multi-scale channel attention mechanism and multi-scale spatial attention mechanism. Subsequently, we employed adaptive hierarchical weighting processes on the extracted building features. Concurrently, we introduced a gating mechanism to facilitate the effective fusion of multi-scale features. The efficacy of the proposed MSANet was evaluated using the WHU aerial image dataset and the WHU satellite image dataset. The experimental results demonstrate compelling performance metrics, with the F1 scores registering at 93.76% and 77.64% on the WHU aerial imagery dataset and WHU satellite dataset II, respectively. Furthermore, the intersection over union (IoU) values stood at 88.25% and 63.46%, surpassing benchmarks set by DeepLabV3 and GSMC.

In remote sensing object detection, the extreme scale variations of objects and the object truncation caused by tiling processing have persistently limited objects localization accuracy and reliability. To effectively address these challenges, this paper proposes a novel localization-oriented detection framework for precise object localization and boundary recovery. First, we enhance feature representation of local details and global context through the effective collaboration of hierarchical attention and coordinate attention mechanisms, enabling accurate capture of spatial and geometric attributes. Second, an adaptive bidirectional attention guidance mechanism is established to achieve mutual reinforcement between high-level semantics and low-level features, improving object localization accuracy and multi-scale adaptability. Finally, by analyzing the spatial geometric relationships between detection boxes and image boundaries, we develop a boundary-aware fusion strategy to effectively identify and merge detection boxes of truncated object, thereby mitigating localization errors induced by tiling. Experiments on DOTA-v1.5, DOTA-v2.0, and DIOR datasets show that our method achieves consistent improvements over representative baselines and improves robustness under object scale variation and tile-boundary truncation, while maintaining a competitive trade-off between detection accuracy and computational cost.