The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
S. No. 
Level-1 
Mode of 
Feature 
Extraction 
Level-2 
1 
Communication 
Line 
2D (Cl & 
C2) 
Road with Trees 
National 
Highway 
City Roads / 
Streets 
Cart Track 
Railway Line 
Rail under 
Construction 
Road Under 
Construction 
2 
Construction 
2D & 3D 
Building / 
Building Blocks 
Houses in Rural 
Area 
Stadium 
Aerodrome 
Railway Station 
Industrial area 
3 
Hydrography 
2D (Cl 
&C2) 
River / Stream 
permanent 
River / Stream 
Temporary 
River with trees 
Canal 
Lake / Pond / 
Water bodies 
Permanent 
Lake / Pond / 
Water bodies 
temporary 
Barrage / 
Reservoir 
4 
Orographie 
3D (Cl) 
Index Contour 
Contour Normal 
5 
Vegetation 
2D/3D 
Vegetation Limit 
Forest 
Plantation / Trees 
Table - 1: List Of features Captured from Cartosat-1 and 
Cartosat-2 Orthoimages in 2D and Cartosat-1 stereo mode in 
3D 
Terrain (DEM) can be accurately extracted from Cartosat-1 
stereo data, while the achieved accuracy of the DEM is 
depending upon the pixel size, the base to height relation, the 
contrast and slope of the area, but also the time interval between 
imaging both scenes. On these parameters of accuracy Cartosat- 
1 fits perfectly. The DEM from the Cartosat-1 data has been 
generated at 10 m grid interval as shown in figure-4. The other 
information content (viz- forest, vegetation, roads, lakes and 
buildings) have been extracted by Cartosat-1 stereo images (3- 
D feature extraction) and further supplemented by Cartosat-2 
(2-D feature extraction) as in figure-5. Older maps, City/district 
maps and Mobile mapping system (Palmtop & hand held GPS) 
are the other sources to collect the geographic names. There is a 
tremendous reduction in the ground truth data collection (field 
verification) by using Cartosat-1 and Cartosat-2 datasets 
together in an integrated manner. The accuracy and 
morphologic information contents in both the satellites are 
sufficient for most of the mapping application at the 1:25000 
scales, which can further be extended to 1:10000 scales (with 
the information content as shown in figure 1 (a)). 
Figure-1(b) Cartosat-1 and Cartosat-2 images at 1 m 
Figure-2(a) Comparison of features extracted in 2D mode (Red: 
Cartosat-1 & Blue: Cartosat-2)