Re-DLinkNet: Based on DLinkNet and ReNet for Road Extraction from High Resolution Satellite Imagery

In order to speed up the update of existing road maps, it is crucial to develop a more efficient road extraction method from remote sensing images. In recent years, deep learning techniques have been widely used for road extraction applications. Among current CNN-based deep networks for road extract...

Full description

Saved in:
Bibliographic Details
Main Authors: Wang, Yuchuan, Tong, Ling, Wen, Jiang, Xiao, Fanghong, Gao, Yaqi, He, Liubei, Li, DingMao
Format: Conference Proceeding
Language:English
Subjects:
Convolution
DenseRe-Block
DLinkNet
Feature extraction
Image Processing
Optical fiber networks
Remote Sensing
Road Extraction
Roads
Satellites
Shape
Urban areas
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 4667
container_issue
container_start_page 4664
container_title
container_volume
creator Wang, Yuchuan
Tong, Ling
Wen, Jiang
Xiao, Fanghong
Gao, Yaqi
He, Liubei
Li, DingMao
description In order to speed up the update of existing road maps, it is crucial to develop a more efficient road extraction method from remote sensing images. In recent years, deep learning techniques have been widely used for road extraction applications. Among current CNN-based deep networks for road extraction, few works study the shape of the convolution kernel, and the remote contextual features dependency relationship is not fully utilized. In view of these problems, an improved DLinkNet is proposed in this paper. Firstly, a convolutional layer which fuses information of multiple scales is used to replace the InitBlock in the front of the network. Secondly, instead of using the D-Block, the DenseRe-Block is applied to the center structure of DLinkNet. The experimental results show that the improved network has a higher IoU score than DLinkNet when extracting roads from optical images in both city and mountain town areas.
doi_str_mv 10.1109/IGARSS47720.2021.9553728
format conference_proceeding
fullrecord <record><control><sourceid>ieee_CHZPO</sourceid><recordid>TN_cdi_ieee_primary_9553728</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9553728</ieee_id><sourcerecordid>9553728</sourcerecordid><originalsourceid>FETCH-LOGICAL-i118t-4e9ccb9d120a27a74ff7c5b6ce6f1039e3605e0442005e1eafe39d00208f486f3</originalsourceid><addsrcrecordid>eNo1kNtKw0AYhFdBsNY-gTf7Aqn_nrPe1dpDICgkel22yb91NQdJIti3N2q9mmH4GIYhhDKYMwb2NtkssjyXxnCYc-BsbpUShsdnZGZNzLRWEoS2-pxMOFMiMgDiklz1_dtoYg4wIUWG0UMamvdHHO7oveuxpG1D_yPqmpJm-ON829GsdSVdfQ2dK4YwYr5ra7oNh9eR6dvq8zfM3YBVFQakSe0O2B2vyYV3VY-zk07Jy3r1vNxG6dMmWS7SKDAWD5FEWxR7WzIOjhtnpPemUHtdoPYMhEWhQSFIOQ5XyNB5FLYE4BB7GWsvpuTmrzcg4u6jC7XrjrvTJ-IbTbBWfQ</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Re-DLinkNet: Based on DLinkNet and ReNet for Road Extraction from High Resolution Satellite Imagery</title><source>IEEE Xplore All Conference Series</source><creator>Wang, Yuchuan ; Tong, Ling ; Wen, Jiang ; Xiao, Fanghong ; Gao, Yaqi ; He, Liubei ; Li, DingMao</creator><creatorcontrib>Wang, Yuchuan ; Tong, Ling ; Wen, Jiang ; Xiao, Fanghong ; Gao, Yaqi ; He, Liubei ; Li, DingMao</creatorcontrib><description>In order to speed up the update of existing road maps, it is crucial to develop a more efficient road extraction method from remote sensing images. In recent years, deep learning techniques have been widely used for road extraction applications. Among current CNN-based deep networks for road extraction, few works study the shape of the convolution kernel, and the remote contextual features dependency relationship is not fully utilized. In view of these problems, an improved DLinkNet is proposed in this paper. Firstly, a convolutional layer which fuses information of multiple scales is used to replace the InitBlock in the front of the network. Secondly, instead of using the D-Block, the DenseRe-Block is applied to the center structure of DLinkNet. The experimental results show that the improved network has a higher IoU score than DLinkNet when extracting roads from optical images in both city and mountain town areas.</description><identifier>EISSN: 2153-7003</identifier><identifier>EISBN: 9781665403696</identifier><identifier>EISBN: 1665403691</identifier><identifier>DOI: 10.1109/IGARSS47720.2021.9553728</identifier><language>eng</language><publisher>IEEE</publisher><subject>Convolution ; DenseRe-Block ; DLinkNet ; Feature extraction ; Image Processing ; Optical fiber networks ; Remote Sensing ; Road Extraction ; Roads ; Satellites ; Shape ; Urban areas</subject><ispartof>2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, 2021, p.4664-4667</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9553728$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,27925,54555,54932</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/9553728$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Wang, Yuchuan</creatorcontrib><creatorcontrib>Tong, Ling</creatorcontrib><creatorcontrib>Wen, Jiang</creatorcontrib><creatorcontrib>Xiao, Fanghong</creatorcontrib><creatorcontrib>Gao, Yaqi</creatorcontrib><creatorcontrib>He, Liubei</creatorcontrib><creatorcontrib>Li, DingMao</creatorcontrib><title>Re-DLinkNet: Based on DLinkNet and ReNet for Road Extraction from High Resolution Satellite Imagery</title><title>2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS</title><addtitle>IGARSS</addtitle><description>In order to speed up the update of existing road maps, it is crucial to develop a more efficient road extraction method from remote sensing images. In recent years, deep learning techniques have been widely used for road extraction applications. Among current CNN-based deep networks for road extraction, few works study the shape of the convolution kernel, and the remote contextual features dependency relationship is not fully utilized. In view of these problems, an improved DLinkNet is proposed in this paper. Firstly, a convolutional layer which fuses information of multiple scales is used to replace the InitBlock in the front of the network. Secondly, instead of using the D-Block, the DenseRe-Block is applied to the center structure of DLinkNet. The experimental results show that the improved network has a higher IoU score than DLinkNet when extracting roads from optical images in both city and mountain town areas.</description><subject>Convolution</subject><subject>DenseRe-Block</subject><subject>DLinkNet</subject><subject>Feature extraction</subject><subject>Image Processing</subject><subject>Optical fiber networks</subject><subject>Remote Sensing</subject><subject>Road Extraction</subject><subject>Roads</subject><subject>Satellites</subject><subject>Shape</subject><subject>Urban areas</subject><issn>2153-7003</issn><isbn>9781665403696</isbn><isbn>1665403691</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2021</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNo1kNtKw0AYhFdBsNY-gTf7Aqn_nrPe1dpDICgkel22yb91NQdJIti3N2q9mmH4GIYhhDKYMwb2NtkssjyXxnCYc-BsbpUShsdnZGZNzLRWEoS2-pxMOFMiMgDiklz1_dtoYg4wIUWG0UMamvdHHO7oveuxpG1D_yPqmpJm-ON829GsdSVdfQ2dK4YwYr5ra7oNh9eR6dvq8zfM3YBVFQakSe0O2B2vyYV3VY-zk07Jy3r1vNxG6dMmWS7SKDAWD5FEWxR7WzIOjhtnpPemUHtdoPYMhEWhQSFIOQ5XyNB5FLYE4BB7GWsvpuTmrzcg4u6jC7XrjrvTJ-IbTbBWfQ</recordid><startdate>20210711</startdate><enddate>20210711</enddate><creator>Wang, Yuchuan</creator><creator>Tong, Ling</creator><creator>Wen, Jiang</creator><creator>Xiao, Fanghong</creator><creator>Gao, Yaqi</creator><creator>He, Liubei</creator><creator>Li, DingMao</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>20210711</creationdate><title>Re-DLinkNet: Based on DLinkNet and ReNet for Road Extraction from High Resolution Satellite Imagery</title><author>Wang, Yuchuan ; Tong, Ling ; Wen, Jiang ; Xiao, Fanghong ; Gao, Yaqi ; He, Liubei ; Li, DingMao</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i118t-4e9ccb9d120a27a74ff7c5b6ce6f1039e3605e0442005e1eafe39d00208f486f3</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Convolution</topic><topic>DenseRe-Block</topic><topic>DLinkNet</topic><topic>Feature extraction</topic><topic>Image Processing</topic><topic>Optical fiber networks</topic><topic>Remote Sensing</topic><topic>Road Extraction</topic><topic>Roads</topic><topic>Satellites</topic><topic>Shape</topic><topic>Urban areas</topic><toplevel>online_resources</toplevel><creatorcontrib>Wang, Yuchuan</creatorcontrib><creatorcontrib>Tong, Ling</creatorcontrib><creatorcontrib>Wen, Jiang</creatorcontrib><creatorcontrib>Xiao, Fanghong</creatorcontrib><creatorcontrib>Gao, Yaqi</creatorcontrib><creatorcontrib>He, Liubei</creatorcontrib><creatorcontrib>Li, DingMao</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Xplore</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Wang, Yuchuan</au><au>Tong, Ling</au><au>Wen, Jiang</au><au>Xiao, Fanghong</au><au>Gao, Yaqi</au><au>He, Liubei</au><au>Li, DingMao</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Re-DLinkNet: Based on DLinkNet and ReNet for Road Extraction from High Resolution Satellite Imagery</atitle><btitle>2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS</btitle><stitle>IGARSS</stitle><date>2021-07-11</date><risdate>2021</risdate><spage>4664</spage><epage>4667</epage><pages>4664-4667</pages><eissn>2153-7003</eissn><eisbn>9781665403696</eisbn><eisbn>1665403691</eisbn><abstract>In order to speed up the update of existing road maps, it is crucial to develop a more efficient road extraction method from remote sensing images. In recent years, deep learning techniques have been widely used for road extraction applications. Among current CNN-based deep networks for road extraction, few works study the shape of the convolution kernel, and the remote contextual features dependency relationship is not fully utilized. In view of these problems, an improved DLinkNet is proposed in this paper. Firstly, a convolutional layer which fuses information of multiple scales is used to replace the InitBlock in the front of the network. Secondly, instead of using the D-Block, the DenseRe-Block is applied to the center structure of DLinkNet. The experimental results show that the improved network has a higher IoU score than DLinkNet when extracting roads from optical images in both city and mountain town areas.</abstract><pub>IEEE</pub><doi>10.1109/IGARSS47720.2021.9553728</doi><tpages>4</tpages></addata></record>
fulltext fulltext_linktorsrc
identifier EISSN: 2153-7003
ispartof 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, 2021, p.4664-4667
issn 2153-7003
language eng
recordid cdi_ieee_primary_9553728
source IEEE Xplore All Conference Series
subjects Convolution
DenseRe-Block
DLinkNet
Feature extraction
Image Processing
Optical fiber networks
Remote Sensing
Road Extraction
Roads
Satellites
Shape
Urban areas
title Re-DLinkNet: Based on DLinkNet and ReNet for Road Extraction from High Resolution Satellite Imagery
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-07T22%3A36%3A11IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_CHZPO&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Re-DLinkNet:%20Based%20on%20DLinkNet%20and%20ReNet%20for%20Road%20Extraction%20from%20High%20Resolution%20Satellite%20Imagery&rft.btitle=2021%20IEEE%20International%20Geoscience%20and%20Remote%20Sensing%20Symposium%20IGARSS&rft.au=Wang,%20Yuchuan&rft.date=2021-07-11&rft.spage=4664&rft.epage=4667&rft.pages=4664-4667&rft.eissn=2153-7003&rft_id=info:doi/10.1109/IGARSS47720.2021.9553728&rft.eisbn=9781665403696&rft.eisbn_list=1665403691&rft_dat=%3Cieee_CHZPO%3E9553728%3C/ieee_CHZPO%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i118t-4e9ccb9d120a27a74ff7c5b6ce6f1039e3605e0442005e1eafe39d00208f486f3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=9553728&rfr_iscdi=true