The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part Bl. Beijing 2008
panchromatic and multispectral data of current digital
photogrammetric cameras pose completely new challenges to
analysis and especially to accuracy requirements on previous
co-registration of image data, which need to be (object-related)
in the centimetre range (see figure 2). For remote sensing tasks
one of the true colour bands is replaced by the N1R band (CIR-
image).
Integration of image and elevation data is in principle resolved
by orthophoto generation in photogrammetry and has been
standardized in Germany (German Institute for Standardization
DIN, 2003). Modem technologies (Laserscanner and InSAR-
systems) also permit the generation of high-resolution DEMs in
complicated observation conditions (e.g. in forests, in urban
areas, in conditions with poor contrast and in bad weather
conditions). However, formal approaches remain the same as
already defined. Merely co-registration becomes more
Figure 2. Panchromatic (top), RGB (middle) and CIR (bottom)
images (image data by DMC/Intergraph)
complicated because of different observation conditions (e.g.
SAR-systems).
Latest key research developments are towards true orthophotos.
Moreover, texture and 3D information can be obtained by
airborne cameras (e.g. MFC). This is applied for 3D city models
from stereo image data, with a remarkable sharpness of the
edges of the 3D objects (Boemer et al., 2008) (see figure 3-4).
Figure 3. Digital Elevation Model (DEM) (image data by MFC)
Methods for resolution enhancement have been established in
research for many years and also employed in production.
Keywords are linked to SPOT HR band and ADS40, which are
equipped with what is known as a “staggered array” (Reulke et
al. 2004) and also with Superresolution (SR).
SR algorithms can be divided in frequency and spatial domain
algorithms. Through SR enhanced-resolution images are created
by merging lower-resolution images together. Reulke et al.
showed an improved resolution by a factor of 2 using a
combination of staggered data and image restoration. A
resolution potential in the sub-decimetre range was
