Proceedings of the KSRS Conference (대한원격탐사학회:학술대회논문집)

  • 2002.10a
  • /
  • Pages.95-100
  • /
  • 2002
  • /
  • 1226-9743(pISSN)
Korean Society of Remote Sensing (대한원격탐사학회)
권호 표지

Detection of buildings from 1m-resolution satellite imagery

  • Kim, Sung-Chai (Dept. of Electronic Engineering, Sogang University) ;
  • Jeon, Seung-Hun (Dept. of Electronic Engineering, Sogang University) ;
  • Kim, Min (Dept. of Electronic Engineering, Sogang University) ;
  • Lee, Kwae-Hi (Dept. of Electronic Engineering, Sogang University)
  • Published : 2002.10.01
Download PDF
⟨ Previous Next ⟩

Abstract

Detecting simple shaped buildings from 1m-resolution satellite imagery is presented. The proposed algorithm is that first, image features such as edges are detected and then segmentation process is performed with the detected features. It can be result in line primitives. These primitives are linked and grouped by building hypotheses. Proposed building hypotheses restrict a building to simple rectangular shape. And sub-region homogeneity test is performed for finding rooftops of buildings. The proposed algorithm has been tested on IKONOS satellite image with 1m-resolution.

Keywords