SnooperText: A text detection system for automatic scenes indexing of urban scenes

dc.creatorMinetto, R
dc.creatorThome, N
dc.creatorCord, M
dc.creatorLeite, NJ
dc.creatorStolfi, J
dc.date2014
dc.dateMAY
dc.date2014-07-30T18:42:59Z
dc.date2015-11-26T17:39:54Z
dc.date2014-07-30T18:42:59Z
dc.date2015-11-26T17:39:54Z
dc.date.accessioned2018-03-29T00:21:30Z
dc.date.available2018-03-29T00:21:30Z
dc.identifierComputer Vision And Image Understanding. Academic Press Inc Elsevier Science, v. 122, n. 92, n. 104, 2014.
dc.identifier1077-3142
dc.identifier1090-235X
dc.identifierWOS:000334394900009
dc.identifier10.1016/j.cviu.2013.10.004
dc.identifierhttp://www.repositorio.unicamp.br/jspui/handle/REPOSIP/71707
dc.identifierhttp://repositorio.unicamp.br/jspui/handle/REPOSIP/71707
dc.identifier.urihttp://repositorioslatinoamericanos.uchile.cl/handle/2250/1286570
dc.descriptionFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
dc.descriptionConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
dc.descriptionCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.descriptionWe describe SNOOPERTEXT, an original detector for textual information embedded in photos of building facades (such as names of stores, products and services) that we developed for the iTowns urban geographic information project. SNOOPERTEXT locates candidate characters by using toggle-mapping image segmentation and character/non-character classification based on shape descriptors. The candidate characters are then grouped to form either candidate words or candidate text lines. These candidate regions are then validated by a text/non-text classifier using a HOG-based descriptor specifically tuned to single-line text regions. These operations are applied at multiple image scales in order to suppress irrelevant detail in character shapes and to avoid the use of overly large kernels in the segmentation. We show that SNOOPERTEXT outperforms other published state-of-the-art text detection algorithms on standard image benchmarks. We also describe two metrics to evaluate the end-to-end performance of text extraction systems, and show that the use of SNOOPERTEXT as a pre-filter significantly improves the performance of a general-purpose OCR algorithm when applied to photos of urban scenes. (C) 2013 Elsevier Inc. All rights reserved.
dc.description122
dc.description92
dc.description104
dc.descriptionFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
dc.descriptionConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
dc.descriptionCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.descriptionANR [07-MDC0-007-03]
dc.descriptionFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
dc.descriptionConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
dc.descriptionCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.descriptionFAPESP [07/54201-6, 07/52015-0]
dc.descriptionCNPq [301016/92-5]
dc.descriptionCAPES [592/08]
dc.descriptionANR [07-MDC0-007-03]
dc.languageen
dc.publisherAcademic Press Inc Elsevier Science
dc.publisherSan Diego
dc.publisherEUA
dc.relationComputer Vision And Image Understanding
dc.relationComput. Vis. Image Underst.
dc.rightsfechado
dc.rightshttp://www.elsevier.com/about/open-access/open-access-policies/article-posting-policy
dc.sourceWeb of Science
dc.subjectText detection
dc.subjectText region classification
dc.subjectHistogram of oriented gradients for text
dc.subjectText descriptor
dc.subjectTextual indexing in urban scene images
dc.subjectImages
dc.subjectExtraction
dc.subjectVideo
dc.subjectClassification
dc.subjectRecognition
dc.titleSnooperText: A text detection system for automatic scenes indexing of urban scenes
dc.typeArtículos de revistas


Este ítem pertenece a la siguiente institución