Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference (한국측량학회:학술대회논문집)
- 2010.04a
- /
- Pages.277-280
- /
- 2010
Automatic Extraction of Land Cover information By Using KOMPSAT-2 Imagery
KOMPSAT-2 영상을 이용한 토지피복정보 자동 추출
- Published : 2010.04.23
Abstract
There is a need to convert the old low- or medium-resolution satellite image-based thematic mapping to the high-resolution satellite image-based mapping of GSD 1m grade or lower. There is also a need to generate middle- or large-scale thematic maps of 1:5,000 or lower. In this study, the DEM and orthoimage is generated with the KOMPSAT-2 stereo image of Yuseong-gu, Daejeon Metropolitan City. By utilizing the orthoimage, automatic extraction experiments of land cover information are generated for buildings, roads and urban areas, raw land(agricultural land), mountains and forests, hydrosphere, grassland, and shadow. The experiment results show that it is possible to classify, in detail, for natural features such as the hydrosphere, mountains and forests, grassland, shadow, and raw land. While artificial features such as roads, buildings, and urban areas can be easily classified with automatic extraction, there are difficulties on detailed classifications along the boundaries. Further research should be performed on the automation methods using the conventional thematic maps and all sorts of geo-spatial information and mapping techniques in order to classify thematic information in detail.