International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004
5.2.4 Analysis of Brightness Differences by Method
Differences of gray scales by sample bands were calculated and
presented to compare images obtained using Wavelet and IHS
method. Two methods also demonstrated differences of red
levels as shown in Figure 12.
Differences of grey values by bands of fusion images created
by Wavelet and IHS method were presented. Pixels showing
differences from 50 to -50 were the most frequently observed.
While significant differences were occasionally observed, it
was found that fusions were generally processed well.
Consistent differences among Red, Green and Blue band
proved the excellent function of a developed fusion program.
6. CONCLUSION
1. While IHS method is easy to apply, it has disadvantage
because it allows only three bands to be applied. And it
replaces intensities corresponding to image sizes, so it's
sharpness tends to be degraded.
2. Wavelet fusion method has different spatial and spectral
resolutions according to steps applied when fusing images. In
overall consideration, it showed the best results, but it is
required to develop systems to deal with various image sizes
without limits to 2".
3. In accordance with Wavelet fusion using a developed fusion
program, the best result was obtained on Step 4. In this step, the
final image size was 32*32. Thus, if applying wavelet fusion
processing for bigger images(1024*1024, 2048*2048), it is
considered that further steps over Step 4 can be carried out.
4. Existing HPF method demonstrated excellent results on non-
coniferous, grey roofs. And PCA method showed remarkable
results on roads and streams, yellow roofs and red roofs. [HS
fusion method applied in this study presented generally good
results on grey roofs, roads and streams. And and Wavelet
fusion method proved overall good results on green color series,
grey roofs, roads and streams.
Acknowledgement
This work was supported by a grant No. R05-2002- 000-01083-0
from Korea Science & Engineering Foundation.
REFERENCE
Chavez, Pat S., Jr., S. C. Sides, and J. A. Anderson, 1991,
Comparison of Three Different Method to Merge Multi-
resolution and multi-spectral Data: Lansat TM and SPOT
panchromatic, Photogrammetric Engineering and Remote
Sensing.
H.H.Kang, J.H.Park, 2001, Image Fusion Watermarking Based
on the discrete Wavelet Transform, Proc. of Int'l symposium on
Advanced Engineering, pp.377-386.
H.H. Yoo, 2002, Analysis of Image Fusion Methods for the
Generation of Multi-resolution Satellite Image Map, Journal of
Korean Society of Civil Engineers, Vol. 22(4-D), pp.823-833.
Garguet-Duport, B., J. Girel, J. M. Chassery, and G. Pautou,
1996, The Use of Multi-resolution Analysis and Wavelet
Transform for Merging SPOT Panchromatic and multispectral
Image Data, Photogrammetric Engineering and Remote
Sensing.
Jensen, J.R, 1995, Introductory Digital Image Processing ; A
Remote Sensing Perspective, Prentice Hall, USA.
Mallat, S.G., 1989, A Theory for Multi-resolution Signal
Decomposition : The Wavelet Representation, IEEE Trans,
Pattern Anal, Machine Intelligence, Vol.IL, No.7, pp.674-693.
Sunar, F., and N. Musaoglu, 1998, Merging multi-resolution
SPOT P and Landsat TM data: the effects and advantage, /nt. J.
Remote Sensing.
Y.H.Kim, 2000, Comparison of Image Merging Techniques for
Generating High Resolution Optical Image Data, Inha Univ.
thesis, pp.823-833.
Bio-
? Lal
KEY V
ABSTR
Monito:
surface
measure
2/POLI
Hemisp
Area In
upon th
Sensors
areas.
RESUN
Le suiv
surface.
Fonctio
ADEOS
dévelop
l'indice
procédu
capteur:
sélectio
The reg
emissio
Earth" :
the unc
to bett:
betweer
atmospl
The exc
largely
surface,
In clim
are repr
which :
(Sato e
requiret
descript
physica
though
grid ce
variabil
of radia
and Piti
space-b
