Abstract
Deep convolutional neural networks have promoted significant progress in building extraction from high-resolution remote sensing imagery. Although most of such work focuses on modifying existing image segmentation networks in computer vision, we propose a new network in this paper, Deep Encoding Network (DE-Net), that is designed for the very problem based on many lately introduced techniques in image segmentation. Four modules are used to construct DE-Net: the inception-style downsampling modules combining a striding convolution layer and a max-pooling layer, the encoding modules comprising six linear residual blocks with a scaled exponential linear unit (SELU) activation function, the compressing modules reducing the feature channels, and a densely upsampling module that enables the network to encode spatial information inside feature maps. Thus, DE-Net achieves state-of-the-art performance on the WHU Building Dataset in recall, F1-Score, and intersection over union (IoU) metrics without pre-training. It also outperformed several segmentation networks in our self-built Suzhou Satellite Building Dataset. The experimental results validate the effectiveness of DE-Net on building extraction from aerial imagery and satellite imagery. It also suggests that given enough training data, designing and training a network from scratch may excel fine-tuning models pre-trained on datasets unrelated to building extraction.
Highlights
Buildings are fundamental elements of a physical urban environment [1]
Deep Encoding Network (DE-Net) is trained by dice and binary cross-entropy loss to address the sample imbalance problem in building extraction
Note that SRI-Net [47] is directly trained on the WHU dataset, and SRI-Net-UC is pre-trained on the University of California (UC) Merced Land Use Dataset and trained on the WHU dataset
Summary
Buildings are fundamental elements of a physical urban environment [1] Information such as the location, size, and number, of buildings is indispensable for many geographic and social applications, e.g., building thematic mapping [2], land-use mapping [3], urban planning [4], change detection [5], population estimation [6], etc. Such applications require a widely covered range, high accuracy, and regular updates, which makes high-resolution remote sensing (HRRS) imagery the most suitable data source. A desirable method is one that effectively generalizes to various unseen situations, and this is where deep convolution neural networks (DCNNs) shine
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: 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.