Segmentation Is Not the End of Road Extraction: An All-Visible Denoising Autoencoder for Connected and Smooth Road Reconstruction
With a plethora of remote sensing (RS) images, deep neural network-based semantic segmentation model (SegModel) achieves commendable road extraction performance. However, the occlusions caused by vehicles, roadside objects, and shadows cannot be directly identified as road pixels, especially on high...
Uloženo v:
| Vydáno v: | IEEE transactions on geoscience and remote sensing Ročník 61; s. 1 - 18 |
|---|---|
| Hlavní autoři: | , , , , , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
New York
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
2023
|
| Témata: | |
| ISSN: | 0196-2892, 1558-0644 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | With a plethora of remote sensing (RS) images, deep neural network-based semantic segmentation model (SegModel) achieves commendable road extraction performance. However, the occlusions caused by vehicles, roadside objects, and shadows cannot be directly identified as road pixels, especially on high-resolution RS images. Therefore, relying only on a single SegModel to guarantee road connectivity and boundary smoothness in road extraction tasks is extremely difficult. To address this issue, this article puts forward a “segmentation-with-reconstruction” framework, which comprises a SegModel to generate the binary road labels from RS images and a reconstruction model to refine the road labels. Specifically, the former can be compatible with arbitrary existing SegModels, while the latter is built by our proposed model named all-visible denoising autoencoder (AV-DAE). The AV-DAE is designed to be an encoder–decoder architecture that takes topology-corruption road labels as inputs and true road labels as outputs. To better train the AV-DAE, we further present three noise-adding strategies to corrupt road labels for diverse patterns and train the AV-DAE to reconstruct them. Being RS-image-agnostic, the AV-DAE pays more attention to the spatial features rather than the spectral features, which enables it to recover the road topology by improving the connectivity and boundary smoothness. Finally, elaborate simulation results demonstrate that the proposed framework can significantly improve the connectivity and boundary smoothness of the extracted roads while achieving a competitive road extraction performance and high generalization ability compared to the benchmarks. |
|---|---|
| AbstractList | With a plethora of remote sensing (RS) images, deep neural network-based semantic segmentation model (SegModel) achieves commendable road extraction performance. However, the occlusions caused by vehicles, roadside objects, and shadows cannot be directly identified as road pixels, especially on high-resolution RS images. Therefore, relying only on a single SegModel to guarantee road connectivity and boundary smoothness in road extraction tasks is extremely difficult. To address this issue, this article puts forward a “segmentation-with-reconstruction” framework, which comprises a SegModel to generate the binary road labels from RS images and a reconstruction model to refine the road labels. Specifically, the former can be compatible with arbitrary existing SegModels, while the latter is built by our proposed model named all-visible denoising autoencoder (AV-DAE). The AV-DAE is designed to be an encoder–decoder architecture that takes topology-corruption road labels as inputs and true road labels as outputs. To better train the AV-DAE, we further present three noise-adding strategies to corrupt road labels for diverse patterns and train the AV-DAE to reconstruct them. Being RS-image-agnostic, the AV-DAE pays more attention to the spatial features rather than the spectral features, which enables it to recover the road topology by improving the connectivity and boundary smoothness. Finally, elaborate simulation results demonstrate that the proposed framework can significantly improve the connectivity and boundary smoothness of the extracted roads while achieving a competitive road extraction performance and high generalization ability compared to the benchmarks. |
| Author | Zheng, Xiangxiang Ding, Ziyue Han, Lingyi Hou, Lu Zheng, Kan Yang, Haojun |
| Author_xml | – sequence: 1 givenname: Lingyi orcidid: 0000-0003-4202-6260 surname: Han fullname: Han, Lingyi organization: Big Data Center, China Aero Geophysical Survey and Remote Sensing Center for Natural Resources (AGRS), Beijing, China – sequence: 2 givenname: Lu orcidid: 0000-0003-3085-9353 surname: Hou fullname: Hou, Lu organization: Intelligent Computing and Communications (IC2) Laboratory, Wireless Signal Processing and Network (WSPN) Laboratory, Key Laboratory of Universal Wireless Communications, Ministry of Education, Beijing University of Posts and Telecommunications, Beijing, China – sequence: 3 givenname: Xiangxiang surname: Zheng fullname: Zheng, Xiangxiang organization: Big Data Center, China Aero Geophysical Survey and Remote Sensing Center for Natural Resources (AGRS), Beijing, China – sequence: 4 givenname: Ziyue surname: Ding fullname: Ding, Ziyue organization: Big Data Center, China Aero Geophysical Survey and Remote Sensing Center for Natural Resources (AGRS), Beijing, China – sequence: 5 givenname: Haojun orcidid: 0000-0001-7404-5007 surname: Yang fullname: Yang, Haojun organization: Department of Electrical and Computer Engineering, University of Waterloo, Waterloo, ON, Canada – sequence: 6 givenname: Kan orcidid: 0000-0002-8531-6762 surname: Zheng fullname: Zheng, Kan organization: College of Electrical Engineering and Computer Sciences, Ningbo University, Ningbo, China |
| BookMark | eNp9kDtPwzAUhS0EEuXxA9gsMaf4ETsxW1UKVKpAagtr5Dg3EJTaYDsSjPxzXNqJgenc4ZzvSt8JOrTOAkIXlIwpJepqfbdcjRlhfMxZIYWiB2hEhSgzIvP8EI0IVTJjpWLH6CSEN0JoLmgxQt8reNmAjTp2zuJ5wA8u4vgKeGYb7Fq8dLrBs8_otdk2rvHE4knfZ89d6Ooe8A1Yl077gidDdGCNa8Dj1nk8ddaCidBgnVCrjXPxdYdbgnE2RD_8Is_QUav7AOf7PEVPt7P19D5bPN7Np5NFZljBY2Z4XdQ8b4hiNdMKuCStLBjjUBAhFBdlSsIVE0xLIrSWjJKayxqUoNKU_BRd7rjv3n0MEGL15gZv08uKlakry2QvtYpdy3gXgoe2Mt1OTlLQ9RUl1dZ3tfVdbX1Xe99pSf8s33230f7rn80P7w2D7g |
| CitedBy_id | crossref_primary_10_1016_j_rsase_2024_101160 crossref_primary_10_1117_1_JRS_18_024504 crossref_primary_10_3390_s24051708 crossref_primary_10_1109_TIM_2025_3541813 |
| Cites_doi | 10.1016/j.jtte.2016.05.005 10.3390/rs10091461 10.1109/CVPR.2016.278 10.1109/TGRS.2019.2926397 10.1109/36.992826 10.1109/36.763290 10.1109/WACV45572.2020.9093593 10.1109/TGRS.2022.3165817 10.1109/CVPRW.2018.00035 10.3390/ijgi10050329 10.1109/VCIP.2017.8305148 10.1016/j.neucom.2016.03.095 10.1109/ICCV.2019.00502 10.1109/TGRS.2021.3104032 10.1109/TGRS.2021.3128033 10.1109/ICPR48806.2021.9412054 10.3390/rs11091015 10.1109/LGRS.2020.2985774 10.1109/JSTARS.2020.3042816 10.3390/s20072064 10.1007/s12524-017-0702-x 10.1117/12.754563 10.1109/CVPRW.2018.00031 10.1109/LGRS.2017.2704120 10.1109/TGRS.2021.3073923 10.1109/KESE.2009.60 10.1109/TGRS.2010.2041783 10.1109/TGRS.2019.2912301 10.1109/TGRS.2012.2190078 10.1109/JSTARS.2021.3053603 10.1155/2015/784504 10.1109/LGRS.2018.2802944 10.1109/TIP.2021.3117076 10.1080/10798587.2008.10643306 10.1109/TGRS.2020.2964675 10.1109/TGRS.2018.2870871 10.1109/TGRS.2016.2626378 10.1109/TGRS.2022.3197546 10.1109/IGARSS.2011.6049176 10.1109/TGRS.2020.2991733 10.1109/CVPRW.2018.00034 10.1109/JSTARS.2012.2189099 10.1016/j.asoc.2018.05.018 10.3390/rs12193175 10.1109/TGRS.2011.2136381 10.1007/s12524-017-0678-6 10.1109/JSTARS.2020.3036602 10.12677/GST.2015.32005 10.3390/rs11050552 10.1038/nature14539 10.1109/CVPR.2015.7298965 10.3390/rs11060696 10.1109/ICIIP.2011.6108839 10.1109/CVPRW.2018.00036 10.3390/rs12182866 10.1109/TGRS.2020.3016086 10.1109/TGRS.2022.3143855 10.1109/CGiV.2016.27 10.1016/j.isprsjprs.2021.03.016 10.1007/s10462-020-09854-1 10.1155/2019/2373798 10.1109/TGRS.2017.2669341 10.1109/IGARSS.2016.7729406 10.1109/CVPR.2018.00331 10.1109/TGRS.2004.841395 |
| ContentType | Journal Article |
| Copyright | Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023 |
| Copyright_xml | – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2023 |
| DBID | AAYXX CITATION 7UA 8FD C1K F1W FR3 H8D H96 KR7 L.G L7M |
| DOI | 10.1109/TGRS.2023.3276591 |
| DatabaseName | CrossRef Water Resources Abstracts Technology Research Database Environmental Sciences and Pollution Management ASFA: Aquatic Sciences and Fisheries Abstracts Engineering Research Database Aerospace Database Aquatic Science & Fisheries Abstracts (ASFA) 2: Ocean Technology, Policy & Non-Living Resources Civil Engineering Abstracts Aquatic Science & Fisheries Abstracts (ASFA) Professional Advanced Technologies Database with Aerospace |
| DatabaseTitle | CrossRef Aerospace Database Civil Engineering Abstracts Aquatic Science & Fisheries Abstracts (ASFA) Professional Aquatic Science & Fisheries Abstracts (ASFA) 2: Ocean Technology, Policy & Non-Living Resources Technology Research Database ASFA: Aquatic Sciences and Fisheries Abstracts Engineering Research Database Advanced Technologies Database with Aerospace Water Resources Abstracts Environmental Sciences and Pollution Management |
| DatabaseTitleList | Aerospace Database |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Engineering Physics |
| EISSN | 1558-0644 |
| EndPage | 18 |
| ExternalDocumentID | 10_1109_TGRS_2023_3276591 |
| GroupedDBID | -~X 0R~ 29I 4.4 5GY 5VS 6IK 97E AAJGR AASAJ AAWTH AAYXX ABQJQ ABVLG ACGFO ACGFS ACIWK ACNCT AENEX AETIX AFRAH AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CITATION CS3 DU5 EBS EJD F5P HZ~ H~9 IBMZZ ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 O9- OCL P2P RIA RIE RNS RXW TAE TN5 VH1 Y6R 7UA 8FD AARMG ABAZT C1K F1W FR3 H8D H96 KR7 L.G L7M |
| ID | FETCH-LOGICAL-c273t-c3b7b34d092b2a9e360f67223e70559358705039252a605aa6210b36be9516c83 |
| ISICitedReferencesCount | 6 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001000313600025&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0196-2892 |
| IngestDate | Mon Jun 30 07:52:30 EDT 2025 Tue Nov 18 21:28:58 EST 2025 Sat Nov 29 03:32:22 EST 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Language | English |
| License | https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037 |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c273t-c3b7b34d092b2a9e360f67223e70559358705039252a605aa6210b36be9516c83 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 |
| ORCID | 0000-0001-7404-5007 0000-0003-4202-6260 0000-0002-8531-6762 0000-0003-3085-9353 |
| PQID | 2821068023 |
| PQPubID | 85465 |
| PageCount | 18 |
| ParticipantIDs | proquest_journals_2821068023 crossref_citationtrail_10_1109_TGRS_2023_3276591 crossref_primary_10_1109_TGRS_2023_3276591 |
| PublicationCentury | 2000 |
| PublicationDate | 2023-00-00 20230101 |
| PublicationDateYYYYMMDD | 2023-01-01 |
| PublicationDate_xml | – year: 2023 text: 2023-00-00 |
| PublicationDecade | 2020 |
| PublicationPlace | New York |
| PublicationPlace_xml | – name: New York |
| PublicationTitle | IEEE transactions on geoscience and remote sensing |
| PublicationYear | 2023 |
| Publisher | The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
| Publisher_xml | – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE) |
| References | ref13 ref57 ref12 ref56 ref15 ref59 ref14 ref58 ref53 ref52 ref55 ref10 ref54 ref17 ref16 ref19 ref18 garcia-garcia (ref4) 2017 ref51 ref50 ref45 ref48 ref47 ref42 ref41 ref43 ref49 vaswani (ref70) 2017 ref7 cai (ref24) 2013; 25 ref9 van etten (ref75) 2018 he (ref11) 2021 ref6 ref5 devlin (ref8) 2018 ref40 ref35 ref34 ref37 ref36 ref31 ref30 ref74 ref33 ref32 ref2 ref1 ref39 ref38 ronneberger (ref44) 2015 lecun (ref3) 2015; 521 ref73 ref72 ref68 chen (ref69) 2018 ref23 ref67 ref26 ref25 ref20 ref64 ref63 ref22 ref66 ref21 ref65 saxe (ref71) 2013 ref28 ref27 ref29 mnih (ref46) 2013 ref60 ref62 ref61 |
| References_xml | – ident: ref2 doi: 10.1016/j.jtte.2016.05.005 – ident: ref34 doi: 10.3390/rs10091461 – ident: ref9 doi: 10.1109/CVPR.2016.278 – ident: ref36 doi: 10.1109/TGRS.2019.2926397 – ident: ref15 doi: 10.1109/36.992826 – ident: ref47 doi: 10.1109/36.763290 – ident: ref42 doi: 10.1109/WACV45572.2020.9093593 – year: 2018 ident: ref75 article-title: SpaceNet: A remote sensing dataset and challenge series publication-title: arXiv 1807 01232 – ident: ref64 doi: 10.1109/TGRS.2022.3165817 – ident: ref39 doi: 10.1109/CVPRW.2018.00035 – ident: ref58 doi: 10.3390/ijgi10050329 – ident: ref45 doi: 10.1109/VCIP.2017.8305148 – ident: ref1 doi: 10.1016/j.neucom.2016.03.095 – ident: ref72 doi: 10.1109/ICCV.2019.00502 – ident: ref61 doi: 10.1109/TGRS.2021.3104032 – ident: ref67 doi: 10.1109/TGRS.2021.3128033 – ident: ref66 doi: 10.1109/ICPR48806.2021.9412054 – ident: ref52 doi: 10.3390/rs11091015 – ident: ref38 doi: 10.1109/LGRS.2020.2985774 – ident: ref53 doi: 10.1109/JSTARS.2020.3042816 – year: 2021 ident: ref11 article-title: Masked autoencoders are scalable vision learners publication-title: arXiv 2111 06377 – ident: ref68 doi: 10.3390/s20072064 – ident: ref21 doi: 10.1007/s12524-017-0702-x – ident: ref23 doi: 10.1117/12.754563 – ident: ref73 doi: 10.1109/CVPRW.2018.00031 – ident: ref26 doi: 10.1109/LGRS.2017.2704120 – ident: ref57 doi: 10.1109/TGRS.2021.3073923 – ident: ref13 doi: 10.1109/KESE.2009.60 – ident: ref17 doi: 10.1109/TGRS.2010.2041783 – ident: ref35 doi: 10.1109/TGRS.2019.2912301 – ident: ref12 doi: 10.1109/TGRS.2012.2190078 – ident: ref37 doi: 10.1109/JSTARS.2021.3053603 – start-page: 6000 year: 2017 ident: ref70 article-title: Attention is all you need publication-title: Proc 31st Int Conf Neural Inf Process Syst – start-page: 801 year: 2018 ident: ref69 article-title: Encoder-decoder with atrous separable convolution for semantic image segmentation publication-title: Proc Eur Conf Comput Vis (ECCV) – ident: ref29 doi: 10.1155/2015/784504 – ident: ref33 doi: 10.1109/LGRS.2018.2802944 – ident: ref56 doi: 10.1109/TIP.2021.3117076 – ident: ref16 doi: 10.1080/10798587.2008.10643306 – ident: ref6 doi: 10.1109/TGRS.2020.2964675 – year: 2013 ident: ref71 article-title: Exact solutions to the nonlinear dynamics of learning in deep linear neural networks publication-title: arXiv 1312 6120 – ident: ref31 doi: 10.1109/TGRS.2018.2870871 – ident: ref18 doi: 10.1109/TGRS.2016.2626378 – ident: ref59 doi: 10.1109/TGRS.2022.3197546 – ident: ref20 doi: 10.1109/IGARSS.2011.6049176 – ident: ref63 doi: 10.1109/TGRS.2020.2991733 – ident: ref50 doi: 10.1109/CVPRW.2018.00034 – ident: ref48 doi: 10.1109/JSTARS.2012.2189099 – start-page: 234 year: 2015 ident: ref44 article-title: U-Net: Convolutional networks for biomedical image segmentation publication-title: Proc Int Conf Med Image Comput Comput -Assist Intervent – ident: ref5 doi: 10.1016/j.asoc.2018.05.018 – ident: ref54 doi: 10.3390/rs12193175 – ident: ref19 doi: 10.1109/TGRS.2011.2136381 – ident: ref27 doi: 10.1007/s12524-017-0678-6 – ident: ref10 doi: 10.1109/JSTARS.2020.3036602 – year: 2013 ident: ref46 article-title: Machine learning for aerial image labeling – year: 2017 ident: ref4 article-title: A review on deep learning techniques applied to semantic segmentation publication-title: arXiv 1704 06857 – ident: ref25 doi: 10.12677/GST.2015.32005 – ident: ref65 doi: 10.3390/rs11050552 – volume: 521 start-page: 436 year: 2015 ident: ref3 article-title: Deep learning publication-title: Nature doi: 10.1038/nature14539 – ident: ref43 doi: 10.1109/CVPR.2015.7298965 – ident: ref41 doi: 10.3390/rs11060696 – ident: ref28 doi: 10.1109/ICIIP.2011.6108839 – ident: ref40 doi: 10.1109/CVPRW.2018.00036 – ident: ref55 doi: 10.3390/rs12182866 – ident: ref60 doi: 10.1109/TGRS.2020.3016086 – ident: ref49 doi: 10.1109/TGRS.2022.3143855 – ident: ref14 doi: 10.1109/CGiV.2016.27 – ident: ref74 doi: 10.1016/j.isprsjprs.2021.03.016 – year: 2018 ident: ref8 article-title: BERT: Pre-training of deep bidirectional transformers for language understanding publication-title: arXiv 1810 04805 – ident: ref7 doi: 10.1007/s10462-020-09854-1 – ident: ref51 doi: 10.1155/2019/2373798 – ident: ref62 doi: 10.1109/TGRS.2017.2669341 – ident: ref30 doi: 10.1109/IGARSS.2016.7729406 – volume: 25 start-page: 25 year: 2013 ident: ref24 article-title: Optimized method for road extraction from high resolution remote sensing image based on watershed algorithm publication-title: Remote Sens Land Resour – ident: ref32 doi: 10.1109/CVPR.2018.00331 – ident: ref22 doi: 10.1109/TGRS.2004.841395 |
| SSID | ssj0014517 |
| Score | 2.4499092 |
| Snippet | With a plethora of remote sensing (RS) images, deep neural network-based semantic segmentation model (SegModel) achieves commendable road extraction... |
| SourceID | proquest crossref |
| SourceType | Aggregation Database Enrichment Source Index Database |
| StartPage | 1 |
| SubjectTerms | Artificial neural networks Benchmarks Coders Connectivity Corruption Image processing Image reconstruction Image resolution Image segmentation Labels Neural networks Noise reduction Remote sensing Roads & highways Roadsides Semantic segmentation Smoothness Topology |
| Title | Segmentation Is Not the End of Road Extraction: An All-Visible Denoising Autoencoder for Connected and Smooth Road Reconstruction |
| URI | https://www.proquest.com/docview/2821068023 |
| Volume | 61 |
| WOSCitedRecordID | wos001000313600025&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 1558-0644 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014517 issn: 0196-2892 databaseCode: RIE dateStart: 19800101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV3fb9MwELbKAIk9IBggBgP5gSeqjPx0Yt4qVGASmhArU8VL5DhOFSlLpqWdusf9c_xd3NlOmjKB2AMvVmu1lzT39Xx3vvtMyBsIcAJfFJ5TCFk4YS6YkzDJnUAJJvPAlYUu-T_9Eh8fJ_M5_zoa_ex6YS6ruK6T9Zqf_1dVwxwoG1tnb6HuXihMwGtQOoygdhj_SfEnanFmG4rq8RGYsUa3K46nda6ZRxqRj6fr5YXpaLCJwUlVOacl_D0qrCOqm1KnECarZYNEl8g3geWIuipGoo-qCz7PGlCzEYhR7IaLdujxYjSJB1F0p5Lr7YmFarp-Il3frgAwatxiMb1dSbVRrG3WYHFV9pPNSk-uNhlvZczVHIC-WOPQO-f2vJYf5dVKDdMbpve4y3Vy5kA8aIy1svY5ShzwosKhATds7tYCe4Ol3Fj2m4uE5lidffp2cohXPAz8mEXc26yIXRXAbwtlX76oAyeXpygiRRGpFXGH3PXjiCemj7DfzQojz7btmx9kd9dBxLsbd7HtH227B9rnmT0iD22wQicGZI_JSNV7ZHdAYblH7usSYtk-IddD4NGjlgLwKACPAvBoU1DECd0A7z2d1HQAO9rDjg5gRwF2tIcdBbBQAzsjbht2T8n3j9PZh8-OPeDDkeA1Lx0ZZHEWhLnL_cwXXAXMLVgMDqtCjifcoY-Rroj7kS8g7BaC-Z6bBSxTEBcwmQTPyE7d1Oo5oUkU-nnoZS7PsjASXqIit4hEwaJCCjdj-8TtHmsqLfs9HsJSpX9U5j5523_l3FC__O3DB52uUmsM2tRP4H4ZUiy-uI2sl-QBvjWJvgOyA09RvSL35OWybC9ea2j9AvjYr-k |
| linkProvider | IEEE |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Segmentation+Is+Not+the+End+of+Road+Extraction%3A+An+All-Visible+Denoising+Autoencoder+for+Connected+and+Smooth+Road+Reconstruction&rft.jtitle=IEEE+transactions+on+geoscience+and+remote+sensing&rft.au=Han%2C+Lingyi&rft.au=Hou%2C+Lu&rft.au=Zheng%2C+Xiangxiang&rft.au=Ding%2C+Ziyue&rft.date=2023&rft.issn=0196-2892&rft.eissn=1558-0644&rft.volume=61&rft.spage=1&rft.epage=18&rft_id=info:doi/10.1109%2FTGRS.2023.3276591&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TGRS_2023_3276591 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0196-2892&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0196-2892&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0196-2892&client=summon |