One of the most promising approaches to solve the challenging task of urban land-use classification is integration of multi-spectral satellite image data and range data. In this paper, height data acquired by airborne laser scanning is not only applied for geometric correction of the multi-spectral satellite image data by the generation of ortho-images, but also integrated in urban environment classification. Laser scanning data acquired by Airborne 3D Imager is used to obtain information on the location height above the terrain surface for each pixel. This information can be applied in order to separate urban objects higher than the ground level from objects that are at ground level, e.g., buildings, trees, streets, grass-covered areas, water bodies and bare land etc. It is demonstrated that the classification of urban scenes is significantly improved by integrating multi-spectral and geometric datasets.
[1]
Emmanuel P. Baltsavias,et al.
Airborne laser scanning: existing systems and firms and other resources
,
1999
.
[2]
V. Shettigara,et al.
HEIGHT DETERMINATION OF EXTENDED OBJECTS USING SHADOWS IN SPOT IMAGES
,
1998
.
[3]
H.-P. Bähr,et al.
Wavelet Compression and the Automatic Classification of Urban Environments Using High Resolution Multispectral Imagery and Laser Scanning Data
,
2001,
GeoInformatica.
[4]
Li Shukai,et al.
Data processing technology of airborne 3D image
,
2001
.
[5]
Claus Brenner,et al.
Extraction of buildings and trees in urban environments
,
1999
.