From Detailed Digital Surface Models to City Models Using Constrained Simplification

In: Photogrammetrie, Fernerkundung, Geoinformation (2008):3(207-215)
 

Abstract

We present a method to simplify high-detail full-featured digital surface models (DSM) of cities (i.e. containing the heights of trees, cars, buildings, etc.) geometrically in such a way that all relevant features are preserved, whereas noise and superfluous details collapse. The relevance of features is automatically evaluated using a semantically motivated shape detection and serves as constraint during the simplification. Our results show that we are able to preserve fine details of complex roof structures while all irrelevant features are effectively removed. Thus, we achieve an excellent abstraction of the city data without any interaction of the user, which is not only beneficial for visualization, but could also be used for GIS related applications.

Bilder

Paper herunterladen

Paper herunterladen

Bibtex

@ARTICLE{wahl-2008-detailed,
    author = {Wahl, Roland and Schnabel, Ruwen and Klein, Reinhard},
     pages = {207--215},
     title = {From Detailed Digital Surface Models to City Models Using Constrained Simplification},
   journal = {Photogrammetrie, Fernerkundung, Geoinformation},
    number = {3},
      year = {2008},
  abstract = {We present a method to simplify high-detail full-featured digital surface models (DSM)
              of cities (i.e. containing the heights of trees, cars, buildings, etc.) geometrically in such a way
              that all
              relevant features are preserved, whereas noise and superfluous details collapse. The relevance of
              features is automatically evaluated using a semantically motivated shape detection and serves as
              constraint during the simplification. Our results show that we are able to preserve fine details of
              complex roof structures while all irrelevant features are effectively removed. Thus, we achieve an
              excellent abstraction of the city data without any interaction of the user, which is not only
              beneficial for visualization, but could also be used for GIS related applications.}
}