File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: DE-Net: Deep encoding network for building extraction from high-resolution remote sensing imagery

TitleDE-Net: Deep encoding network for building extraction from high-resolution remote sensing imagery
Authors
KeywordsBuilding extraction
Deep learning
Fully convolutional network
High-resolution remote sensing imagery
Issue Date2019
Citation
Remote Sensing, 2019, v. 11, n. 20, article no. 2380 How to Cite?
AbstractDeep 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.
Persistent Identifierhttp://hdl.handle.net/10722/329585
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorLiu, Hao-
dc.contributor.authorLuo, Jiancheng-
dc.contributor.authorHuang, Bo-
dc.contributor.authorHu, Xiaodong-
dc.contributor.authorSun, Yingwei-
dc.contributor.authorYang, Yingpin-
dc.contributor.authorXu, Nan-
dc.contributor.authorZhou, Nan-
dc.date.accessioned2023-08-09T03:33:51Z-
dc.date.available2023-08-09T03:33:51Z-
dc.date.issued2019-
dc.identifier.citationRemote Sensing, 2019, v. 11, n. 20, article no. 2380-
dc.identifier.urihttp://hdl.handle.net/10722/329585-
dc.description.abstractDeep 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.-
dc.languageeng-
dc.relation.ispartofRemote Sensing-
dc.subjectBuilding extraction-
dc.subjectDeep learning-
dc.subjectFully convolutional network-
dc.subjectHigh-resolution remote sensing imagery-
dc.titleDE-Net: Deep encoding network for building extraction from high-resolution remote sensing imagery-
dc.typeArticle-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.3390/rs11202380-
dc.identifier.scopuseid_2-s2.0-85074187425-
dc.identifier.volume11-
dc.identifier.issue20-
dc.identifier.spagearticle no. 2380-
dc.identifier.epagearticle no. 2380-
dc.identifier.eissn2072-4292-
dc.identifier.isiWOS:000498395800048-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats