A Method for Extracting High-Resolution Building Height Information in Rural Areas Using GF-7 Data.

Building height is important information in disaster management and damage assessment. It is also a key parameter in studies such as population modeling and urbanization. Relatively few studies have been conducted on extracting building height in rural areas using imagery from China's Gaofen-7 satellite (GF-7). In this study, we developed a method combining photogrammetry and deep learning to extract building height using GF-7 data in the rural area of Pingquan in northern China. The deep learning model DELaMa was proposed for digital surface model (DSM) editing based on the Large Mask Inpainting (LaMa) architecture. It not only preserves topographic details but also reasonably predicts the topography inside the building mask. The percentile value of the normalized digital surface model (nDSM) in the building footprint was taken as the building height. The extracted building heights in the study area are highly consistent with the reference building heights measured from the ICESat-2 LiDAR point cloud, with an R2 of 0.83, an MAE of 1.81 m and an RMSE of 2.13 m for all validation buildings. Overall, the proposed method in this paper helps to promote the use of satellite data in large-scale building height surveys, especially in rural areas.

DMU-Net: A Dual-Stream Multi-Scale U-Net Network Using Multi-Dimensional Spatial Information for Urban Building Extraction.

Automatically extracting urban buildings from remote sensing images has essential application value, such as urban planning and management. Gaofen-7 (GF-7) provides multi-perspective and multispectral satellite images, which can obtain three-dimensional spatial information. Previous studies on building extraction often ignored information outside the red-green-blue (RGB) bands. To utilize the multi-dimensional spatial information of GF-7, we propose a dual-stream multi-scale network (DMU-Net) for urban building extraction. DMU-Net is based on U-Net, and the encoder is designed as the dual-stream CNN structure, which inputs RGB images, near-infrared (NIR), and normalized digital surface model (nDSM) fusion images, respectively. In addition, the improved FPN (IFPN) structure is integrated into the decoder. It enables DMU-Net to fuse different band features and multi-scale features of images effectively. This new method is tested with the study area within the Fourth Ring Road in Beijing, and the conclusions are as follows: (1) Our network achieves an overall accuracy (OA) of 96.16% and an intersection-over-union (IoU) of 84.49% for the GF-7 self-annotated building dataset, outperforms other state-of-the-art (SOTA) models. (2) Three-dimensional information significantly improved the accuracy of building extraction. Compared with RGB and RGB + NIR, the IoU increased by 7.61% and 3.19% after using nDSM data, respectively. (3) DMU-Net is superior to SMU-Net, DU-Net, and IEU-Net. The IoU is improved by 0.74%, 0.55%, and 1.65%, respectively, indicating the superiority of the dual-stream CNN structure and the IFPN structure.