Logo: to the web site of Uppsala University

uu.sePublications from Uppsala University
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Segmentation of shadows and water bodies in high resolution images using ancillary data
Akademi för Teknik och Miljö, Högskolan i Gävle.
Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Division of Visual Information and Interaction. Uppsala University, Disciplinary Domain of Science and Technology, Mathematics and Computer Science, Department of Information Technology, Computerized Image Analysis and Human-Computer Interaction.ORCID iD: 0000-0003-0085-5829
2016 (English)In: Proc. 16th International Multidisciplinary Scientific GeoConference, 2016, Vol. 1, p. 827-834Conference paper, Published paper (Refereed)
Abstract [en]

High spatial resolution imagery is often affected by shadows, both in urban environments with large variations in surface elevation and in vegetated areas. It is a common bias in classification when waters and shadows are registered as the same area. The radiometric response for the shadowed regions should be restored prior to classification. To enable that, separate classes of non-shadowed regions and shadowed areas should be created. Previous work on water extraction using low/medium resolution images, mainly faced two difficulties. Firstly, it is difficult to obtain accurate position of water boundary and secondly, shadows of elevated objects e.g. buildings, bridges, towers and trees are a typical source of noise when facing water extraction in urban regions. In high resolution images the problem of separation water and shadows becomes more prominent since the small local variation of intensity values gives rise to misclassification. This paper proposes a robust method for separation of shadowed areas and water bodies in high spatial resolution imagery using hierarchical method on different scales combined with classification of PCA (Principal Component Analysis) bands, which reduces the effects of radiometric and spatial differences that is commonly associated with the pixel-based methods for multisource data fusion. The method uses ancillary data to aid in classification of shadows and waters. The proposed method includes three steps: segmentation, classification and postprocessing. To achieve robust segmentation, we apply the merging region with three features (PCA bands, NSVDI (Normalized Saturation-value Difference Index) and height data). NSVDI discriminates shadows and some water. In the second step we use hierarchic region based classification to identify water regions. After that step candidates for water pixels are verified by the LiDAR DEM data. As a last step we consider shape parameters such as compactness and symmetry to completely remove shadows.

Place, publisher, year, edition, pages
2016. Vol. 1, p. 827-834
Series
International Multidisciplinary Scientific GeoConference-SGEM, ISSN 1314-2704
Keywords [en]
shadows, water bodies, objects, segmentation, high-resolution
National Category
Computer and Information Sciences
Research subject
Computerized Image Processing
Identifiers
URN: urn:nbn:se:uu:diva-309555DOI: 10.5593/SGEM2016/B21/S08.104ISI: 000395499400104ISBN: 978-619-7105-58-2 (print)OAI: oai:DiVA.org:uu-309555DiVA, id: diva2:1052092
Conference
SGEM 2016, June 28 – July 6, Albena Resort, Bulgaria
Available from: 2016-12-05 Created: 2016-12-05 Last updated: 2018-01-13Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full text

Authority records

Seipel, Stefan

Search in DiVA

By author/editor
Seipel, Stefan
By organisation
Division of Visual Information and InteractionComputerized Image Analysis and Human-Computer Interaction
Computer and Information Sciences

Search outside of DiVA

GoogleGoogle Scholar

doi
isbn
urn-nbn

Altmetric score

doi
isbn
urn-nbn
Total: 385 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf