77?e International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B7. Beijing 2008
photographs were taken in 1996 by Zeiss RMK TOP 15 camera
(focal length: 153 mm) with 1:35 000 scale. They are black and
white photographs and were scanned in 21 microns and
orthorectified with 1 meter resolution. Landsat imagery is dated
2000. 28.5 meter resolution 1 st , 2 nd , 3 rd and 14.25 meter
resolution panchromatic bands of the imagery are used for
study. Study area and data are shown in Figure 1.
2. RESOLUTION MERGE OF 1:35.000 SCALE AERIAL
PHOTOGRAPHS WITH LANDSAT 7 ETM IMAGERY
2.1 Direct Merge
In this method, aerial orthophoto is directly merged with
Landsat multispectral imagery. The efficiency of three different
merging techniques (Principal Component, IHS, and Brovey
Transform) is examined. The merged images are shown in
Figure 2 and also a part of imagery is zoomed in and shown in
Figure 3.
Figure 2. Merged imageries (a: Brovey Transform, b: IHS, c:
Principal Component, d: Landsat multispectral)
Figure 3. Merged imageries (a: Brovey Transform, b: IHS, c:
Principal Component, d: Landsat multispectral)
When the imageries examined, it can be easily seen that
Principal Component and Brovey Transform techniques gives
better results. Especially there is a big degradation at the
imagery generated with the IHS technique. Brovey Transform
method creates a more sharpened imagery but with some color
spot areas. Sharpness is less at Principal Component method if
compared with Brovey Transform but no color spots happens in
this method. If the pan-sharpening results from Brovey
Transform, IHS (Intensity-Hue-Saturation) and PC (Principal
Component) are compared, respectively consistent, weak and
superior spectral recovery was achieved with the methods.
2.2 Merge with Resolution Merged Landsat Imagery
In this method, firstly the multispectral Landsat 7 ETM imagery
was merged with Landsat 7 ETM panchromatic imagery and
finaly this imagery was merged with aerial photographs again
with three different merging techniques. The results showed
that this method creates better merged imagery. The merged
images are shown in Figure 4 and also a part of imagery is
zoomed in and shown in Figure 5.
Figure 4. Merged imageries (a: Brovey Transform, b: IHS, c:
Principal Component, d: Resolution merged Landsat
multispectral)
Figure 5. Merged imageries (a: Brovey Transform, b: IHS, c:
Principal Component, d: Resolution merged Landsat
multispectral)
In this method, The Principal Component method serves well
the purpose of spatial resolution enhancement of the low
1283