Advancing Rural Building Extraction via Diverse Dataset Construction and Model Innovation with Attention and Context Learning

Abstract

Rural building automatic extraction technology is of great significance for rural planning and disaster assessment; however, existing methods face the dilemma of scarce sample data and large regional differences in rural buildings. To solve this problem, this study constructed an image dataset of typical Chinese rural buildings, including nine typical geographical regions, such as the Northeast and North China Plains. Additionally, an improved remote sensing image rural building extraction network called AGSC-Net was designed. Based on an encoder–decoder structure, the model integrates multiple attention gate (AG) modules and a context collaboration network (CC-Net). The AG modules realize focused expression of building-related features through feature selection. The CC-Net module models the global dependency between different building instances, providing complementary localization and scale information to the decoder. By embedding AG and CC-Net modules between the encoder and decoder, the model can capture multiscale semantic information on building features. Experiments show that, compared with other models, AGSC-Net achieved the best quantitative metrics on two rural building datasets, verifying the accuracy of the extraction results. This study provides an effective example for automatic extraction in complex rural scenes and lays the foundation for related monitoring and planning applications.

MAD-UNet: A Multi-Region UAV Remote Sensing Network for Rural Building Extraction

“…To further validate the capability of MAD-UNet in extracting rural buildings, a comparative analysis was conducted against five semantic segmentation networks, including U-Net [ 9 ], PSPNet [ 10 ], DeepLabV3+ [ 15 ], HRNet [ 38 ], ISANet [ 39 ], and AGSCNet [ 18 ]. Brief introductions to these models are illustrated as follows:…”

“…(6) AGSCNet [ 18 ] builds upon the U-Net architecture by integrating multiple attention gate modules and a context collaboration network structure. It has achieved good results in tasks involving the extraction of buildings in various rural areas of China.…”

“…ISANet, designed to reduce the computational burden of attention mechanisms while establishing long-range relationships between pixels, excessively focuses on edge connectivity at the expense of clear delineation between buildings, resulting in boundary adhesion. AGSC-Net's CC-Net adopts atrous rates of (1,6,12,18) within ASPP, which for rural small-sized buildings may be overly large, prompting the model to focus too much on contextual information and thus neglecting the details of small-scale targets, causing omissions.…”

“…Yu et al [ 17 ] introduced an attention gate module added to the skip-connection structure of U-Net, which suppresses irrelevant features and noise in the input image, emphasizing the primary features of buildings. Further developing U-Net, Yu et al [ 18 ] created the CC-Net module to connect the encoder and decoder, capturing essential building features, adding multi-scale semantic information to the network, and using multiple attention gates to model global dependencies between buildings. Wang et al [ 19 ] presented a global feature information perception module that addresses the semantic gap between low-level and high-level features and utilizes a spatial attention mechanism, improving the model’s interpretation of spatial feature information.…”