SEMI-AUTOMATIC 3D BUILDING EXTRACTION FROM HIGH RESOLUTION SATELLITE IMAGES

  • Javzandulam, Tsend-Ayush (Department of Geoinformatic Engineering, Inha University) ;
  • Rhee, Soo-Ahm (Department of Geoinformatic Engineering, Inha University) ;
  • Kim, Tae-Jung (Department of Geoinformatic Engineering, Inha University) ;
  • Kim, Kyung-Ok (Spatial Information Research Team, Electronics and Telecommunications Research Institute)
  • Published : 2006.11.02

Abstract

Extraction of building is one of essential issues for the 3D city models generation. In recent years, high-resolution satellite imagery has become widely available, and this shows an opportunity for the urban mapping. In this paper, we have developed a semi-automatic algorithm to extract 3D buildings in urban settlements areas from high-spatial resolution panchromatic imagery. The proposed algorithm determines building height interactively by projecting shadow regions for a given building height onto image space and by adjusting the building height until the shadow region and actual shadow in the image match. Proposed algorithm is tested with IKONOS images over Deajeon city and the algorithm showed promising results.┌阀؀䭏佈䉌ᔀ鳪떭臬隑駭验耀

Keywords