Abstract
Building extraction from very high resolution (VHR) imagery plays an important role in urban planning, disaster management, navigation, updating geographic databases, and several other geospatial applications. Compared with the traditional building extraction approaches, deep learning networks have recently shown outstanding performance in this task by using both high-level and low-level feature maps. However, it is difficult to utilize different level features rationally with the present deep learning networks. To tackle this problem, a novel network based on DenseNets and the attention mechanism was proposed, called the dense-attention network (DAN). The DAN contains an encoder part and a decoder part which are separately composed of lightweight DenseNets and a spatial attention fusion module. The proposed encoder–decoder architecture can strengthen feature propagation and effectively bring higher-level feature information to suppress the low-level feature and noises. Experimental results based on public international society for photogrammetry and remote sensing (ISPRS) datasets with only red–green–blue (RGB) images demonstrated that the proposed DAN achieved a higher score (96.16% overall accuracy (OA), 92.56% F1 score, 90.56% mean intersection over union (MIOU), less training and response time and higher-quality value) when compared with other deep learning methods.
Highlights
Extracting 2D buildings footprints in very high resolution (VHR) imagery has many applications in navigation, urban planning, disaster management, and population estimation [1]
Based on the used data, building extraction methods can generally be divided into three categories: 2D information based, fused 2D–3D information based and 3D information based [8,9,10]. 2D information is mainly derived from images, including aerial images and space-borne images, while 3D information is mainly derived from airborne laser scanning technology, such as light detection and ranging (LiDAR) data [11,12,13]
A spatial attention fusion module based on the attention mechanism can better guide the low-feature maps to help high-level features recover details of the images and reduce the over-segmentation
Summary
Extracting 2D (two-dimensional) buildings footprints in very high resolution (VHR) imagery has many applications in navigation, urban planning, disaster management, and population estimation [1]. The traditional technologies use handcrafted features as a key feature for building extraction, which may contain spectral information or/and spatial information or/and geometrical information [14]. The performance of these technologies relies on the extraction of low-level hand-engineered local features. The deep learning technologies, as a new framework, have the ability to learn high-level hierarchical features from both 2D/3D information corresponding to the different levels of abstraction, making it dominant in the field of building extraction [15,16]. CNN and FCN only use high-level feature maps to perform pixel-classification; low-level feature maps with rich detailed information are discarded. Some supervised semantic segmentation procedures based on excellent networks such as U-Net [22], DeconvNet [20], Segnet [23], and RefineNet [24] have appeared
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
