Full text: Fusion of sensor data, knowledge sources and algorithms for extraction and classification of topographic objects

International Archives of Photogrammetry and Remote Sensing, Vol. 32, Part 7-4-3 W6, Valladolid, Spain, 3-4 June, 1999 
Original MOMS-2P Band 3 and TM infrared bands 4-5-7, dis 
played as R-G-B colour composite, are shown in Figures 5(a) 
and (b), respectively, together with a version of MOMS-2P spec 
trally enhanced in the infrared wavelength through fusion with 
TM (Fig. 5(c)). Contours and textures are highlighted, since 
both are related to the spatial frequencies injected (see Fig. 5(d)). 
Analogously, original MOMS-2P Band 1 (green), Band 2 (red) 
and TM (visible Bands 3-2-1 displayed as R-G-B true colour) are 
shown in Figure 6(a)-(c), respectively, together with a version of 
MOMS-2P spectrally enhanced in the missing blue wavelength 
through fusion with TM Band 1 (Fig. 6(d)). 
Notice that in both cases spectral signatures -colour hues on 
pictures- of the fused image (Fig. 5(c) and Fig. 6(d)) are dif 
ferent from those of TM (Fig. 5(b) and Fig. 6(c)), since one 
and two bands, respectively, over three straightforwardly come 
from the MOMS-2P sensor which exhibits different responses 
(in gains and offsets) from TM in those spectral bands which 
are common to both. Furthermore, MOMS-2P bands feature also 
spectral responses not perfectly overlapped to those of TM. Un 
like spatial enhancement (Aiazzi, 1998; Chavez, 1991; Nunez 
1999; Wald, 1997), spectral enhancement does imply a global 
change in spectral information with respect to the multi-spectral 
sensor having lower spatial resolution (TM in this case). The 
outcome benefit is that the “enhanced” multi-spectral informa 
tion, which is not a distorted version of that of TM, is available at 
the spatial resolution of the multi-spectral sensor having higher 
ground resolution, i.e. of MOMS-2R Waiting for quantitative 
scores being devised, we may evaluate the quality of fused im 
ages from the sharpness of spatial high-pass features, as well as 
from their matching to contours and textures of the scene (Zhou, 
1998). Therefore, preservation of multi-spectral signatures, al 
though desirable, is on the whole of secondary importance for 
the assessment of the quality of spectral enhancement. Its rel 
evance is, however, primary when, e.g., multi-spectral data are 
being merged with a panchromatic observation (Wald, 1997).

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.