打印

作者

Gabriela Barbosa Martins;Laura Elena Cué La Rosa;Patrick Nigri Happ;Luiz Carlos Teixeira Coelho Filho;Celso Junius F. Santos;Raul Queiroz Feitosa;Matheus Pinheiro Ferreira

来源

URBAN FORESTRY & URBAN GREENING,Vol.64,Issue1,Article 127241

语言

英文

关键字

Tree species discrimination;Deep learning;Convolutional neural networks;Remote sensing

作者单位

Original article"}]},{"#name":"title","$":{"id":"tit0005"},"_":"Deep learning-based tree species mapping in a highly diverse tropical urban setting"}],"floats":[],"footnotes":[],"attachments":[]},"vol-first":"64","vol-iss-suppl-text":"Volume 64","userSettings":{"forceAbstract":false,"creditCardPurchaseAllowed":true,"blockFullTextForAnonymousAccess":false,"disableWholeIssueDownload":false,"preventTransactionalAccess":false,"preventDocumentDelivery":true},"contentType":"JL","crossmark":true,"document-references":50,"freeHtmlGiven":false,"ssoUrls":["//acw.sciencedirect.com/SSOCore/update?acw=52a2962972a0c54ede78d20777eb919d71d9gxrqa%7C%24%7CF2112D79030DFAED63CF6A59A30B470D675AB8F4293FFEE80A7091E8F1D3D6021B41A97B3FADFE804ECF3C010F8EC99E908EAEE75C7A83F03FBA44D1BD4E4F2EB0469A67597464825D387A21AFA2E514&utt=5c567ae9b5dfc718f8c410464d36fd2863ab6b2","//null-null/update?acw=52a2962972a0c54ede78d20777eb919d71d9gxrqa%7C%24%7CF2112D79030DFAED63CF6A59A30B470D675AB8F4293FFEE80A7091E8F1D3D6021B41A97B3FADFE804ECF3C010F8EC99E908EAEE75C7A83F03FBA44D1BD4E4F2EB0469A67597464825D387A21AFA2E514&utt=5c567ae9b5dfc718f8c410464d36fd2863ab6b2"],"userProfile":{"departmentName":"ScienceDirect Guests","accessType":"GUEST","accountId":"228598","webUserId":"12975512","accountName":"ScienceDirect Guests","departmentId":"291352","userType":"NORMAL","hasMultipleOrganizations":false},"access":{"openAccess":false,"openArchive":false},"aipType":"none","articleEntitlement":{"entitled":false,"isCasaUser":false,"usageInfo":"(12975512,U|291352,D|228598,A|3,P|2,PL)(SDFE,CON|52a2962972a0c54ede78d20777eb919d71d9gxrqa,SSO|ANON_GUEST,ACCESS_TYPE)"},"isThirdParty":false,"crawlerInformation":{"canCrawlPDFContent":false,"isCrawler":false},"dates":{"Available online":"2 July 2021","Received":"12 March 2021","Revised":["9 June 2021"],"Accepted":"28 June 2021","Publication date":"1 September 2021"},"displayViewFullText":false,"downloadFullIssue":false,"entitlementReason":"unsubscribed","hasBody":true,"hasScholarlyAbstract":true,"headerConfig":{"helpUrl":"https://service.elsevier.com/app/home/supporthub/sciencedirect/","contactUrl":"https://service.elsevier.com/app/contact/supporthub/sciencedirect/","userName":"","userEmail":"","orgName":"ScienceDirect Guests","webUserId":"12975512","libraryBanner":{},"shib_regUrl":"","tick_regUrl":"","recentInstitutions":[],"canActivatePersonalization":false,"hasInstitutionalAssociation":false,"hasMultiOrg":false,"userType":"GUEST","userAnonymity":"ANON_GUEST","allowCart":true,"environment":"prod","cdnAssetsHost":"https://sdfestaticassets-us-east-1.sciencedirectassets.com"},"isCorpReq":false,"issn":"16188667","issn-primary-formatted":"1618-8667","issRange":"","pageCount":10,"pdfEmbed":false,"publication-content":{"noElsevierLogo":false,"imprintPublisher":{"displayName":"Urban & Fischer","id":"507"},"isSpecialIssue":false,"isSampleIssue":false,"transactionsBlocked":false,"publicationOpenAccess":{"oaStatus":"","oaArticleCount":129,"openArchiveStatus":false,"openArchiveArticleCount":0,"openAccessStartDate":"","oaAllowsAuthorPaid":true},"issue-cover":{"attachment":[{"attachment-eid":"1-s2.0-S1618866721X00076-cov200h.gif","file-basename":"cov200h","extension":"gif","filename":"cov200h.gif","ucs-locator":["https://s3-eu-west-1.amazonaws.com/prod-ucs-content-store-eu-west/content/pii:S1618866721X00076/cover/DOWNSAMPLED200/image/gif/7228e63ff29e30676e4d4749008a3969/cov200h.gif"],"attachment-type":"IMAGE-COVER-H200","filesize":"10591","pixel-height":"200","pixel-width":"150"},{"attachment-eid":"1-s2.0-S1618866721X00076-cov150h.gif","file-basename":"cov150h","extension":"gif","filename":"cov150h.gif","ucs-locator":["https://s3-eu-west-1.amazonaws.com/prod-ucs-content-store-eu-west/content/pii:S1618866721X00076/cover/DOWNSAMPLED/image/gif/12a5959932cde6e42ae86def206fe121/cov150h.gif"],"attachment-type":"IMAGE-COVER-H150","filesize":"7258","pixel-height":"150","pixel-width":"113"}]},"smallCoverUrl":"https://ars.els-cdn.com/content/image/S16188667.gif","title":"urban-forestry-and-urban-greening","contentTypeCode":"JL","sourceOpenAccess":false,"publicationCoverImageUrl":"https://ars.els-cdn.com/content/image/1-s2.0-S1618866721X00076-cov150h.gif"},"useEnhancedReader":true,"volRange":"64","features":["aamAttachments","keywords","references","preview"],"titleString":"Deep learning-based tree species mapping in a highly diverse tropical urban setting","usesAbstractUrl":true,"renderingMode":"Abstract","isAbstract":true,"isContentVisible":false,"ajaxLinks":{"referredToBy":true,"toc":true,"recommendations":true,"citingArticles":true,"authorMetadata":true},"eligibleForUniversalPdf":true},"authors":{"content":[{"#name":"author-group","$":{"id":"aug0005"},"$$":[{"#name":"author","$":{"id":"aut0005","author-id":"S1618866721002661-2983b3962d6ae49663fa28a58f6e3ccf"},"$$":[{"#name":"given-name","_":"Gabriela Barbosa;Cartographic Engineering Department, Military Institute of Engineering (IME), Praça Gen. Tibúrcio 80, 22290-270 Rio de Janeiro, RJ, Brazil;Department of Electrical Engineering, Pontifical Catholic University of Rio de Janeiro (PUC), R. Marquês de São Vicente 225, 22541-041 Rio de Janeiro, RJ, Brazil;Instituto Pereira Passos (IPP), R. Gago Coutinho 52, 22221-070 Rio de Janeiro, RJ, Brazil;Fundação Parques e Jardins, Praça da República, s/n - Campo de Santana, Rio de Janeiro, RJ, Brazil;Cartographic Engineering Department, Military Institute of Engineering (IME), Praça Gen. Tibúrcio 80, 22290-270 Rio de Janeiro, RJ, Brazil;Department of Electrical Engineering, Pontifical Catholic University of Rio de Janeiro (PUC), R. Marquês de São Vicente 225, 22541-041 Rio de Janeiro, RJ, Brazil;Instituto Pereira Passos (IPP), R. Gago Coutinho 52, 22221-070 Rio de Janeiro, RJ, Brazil;Fundação Parques e Jardins, Praça da República, s/n - Campo de Santana, Rio de Janeiro, RJ, Brazil

摘要

Spatially explicit information on urban tree species distribution is crucial for green infrastructure management in cities. This information is usually acquired with ground-based surveys, which are time-consuming and usually cover limited spatial extents. The combination of machine learning algorithms and remote sensing images has been hailed as a promising way to map tree species over broad areas. Recently, convolutional neural networks (CNNs), a type of deep learning method, have achieved outstanding results for tree species discrimination in various remote sensing data types. However, there is a lack of studies using CNN-based methods to produce tree species composition maps, particularly for tropical urban settings. Here, we propose a multi-task CNN to map tree species in a highly diverse neighborhood in Rio de Janeiro, Brazil. Our network architecture takes an aerial photograph (RGB bands and pixel size = 0.15 m) and delivers two outputs: a semantically segmented image and a distance map transform. In the former, all pixel positions are labeled, while in the latter, each pixel position contains the Euclidean distance to the crown boundary. We developed a post-processing approach that combines the two outputs, and we classified nine and five tree species with an average F1-score of 79.3 ± 8.6% and 87.6 ± 4.4%, respectively. Moreover, our post-processing approach produced a realistic tree species composition map by labeling only pixels of the target species with high class membership probabilities. Our results show the potential of CNNs and aerial photographs to map tree species in highly diverse tropical urban settings, providing valuable insights for urban forest management and green spaces planning.