In: Wagner W., Székely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
Since AISA-Dual instrument operated in Israel, has no 
GPS/INS system available, it is incapable to perform the pixel- 
by-pixel geo-location and rectification of the images, generated 
in an extensive HRS dataset. In this study, the advantage of the 
sensors orientation and boresight effects were investigated 
based on AISA-Dual HRS data. Our study shows that those 
"negative" effects are suitable to spectral/spatial analysis and 
processing of HRS images. Thus, we are suggesting three 
efficient and practical applications: 1) enhancing shadowing 
effect, 2) generating a 3-D view, and 3) performing a better 
detection of boarder anomaly. 
2. METHODOLOGY 
Apart from the image sensors, an airborne mobile-mapping 
system has to equip for direct geo-referencing involves one or 
several GPS receivers and antennas as well as an IMU. In the 
most ideal case, all sensors are attached to a common rigid 
mounting structure, preventing variations of their relative 
positions and orientations. In practice, AISA-Dual system is 
operated with no GPS/INS data available. Consequently, it is 
impossible to geo-locate or rectify the VNIR and SWIR images. 
The spatial/spectral boresight effect of AISA-Dual sensor is 
illustrated in Figure 1. 
Figure 1. Boresight effect schematic demonstration, image I and 
II illustrating boresight then I is the target in VNIR wavelengths 
and II is the target in SWIR region, and the spectrum of the 
target then region I is concrete (target in VNIR) and II is asphalt 
(background in SWIR) 
We suggest to converting the boresight shift into additional 
spectral/spatial information by calculating a simple band ratio 
between VNIR and SWIR images. The boresight "band ratio" 
presented in Figure 2. 
Figure 2. Additional Boresight "band" (October 31, 2009; 10:00 
GMT; midlatitude summer model; 28.2 0 solar zenith, 137.6 0 
azimuth angle), A is the 948nm band, B is the lOlOnm band, C 
is the calculated band ratio (948nm/1010nm) interpreted as 
Boresight band 
3. RESULTS 
We have found that the boresight effect have applicable 
outcomes on the spectral/spatial analysis and processing that is 
based on an extensive dataset of AISA-Dual images, acquired 
during more than six years of operated campaigns. Three 
applications were investigated as follow: 1) enhancing 
shadowing effect, 2) generating a 3-D view, and 3) performing a 
better detection of boarder anomaly. 
3.1 Enhancing shadowing effect 
Current implementations of the de-shadowing process 
within the ATCOR-4 model uses image's statistics to gain 
knowledge about the darken area in order to correct the shadow 
effect. This routine is not suitable for data sets acquired on 
clouds shadow as the inter-comparison process is missing in the 
diffuse light conditions. This de-shadowing algorithm consists 
of a sequence of eight processing steps: an atmospheric 
correction, clouds and water bodies masking, five additional 
statistic manipulations including covariance matrixes and 
matched filters to define a core shadow mask, and final step is a 
de-shadowing that exclusively applied to the pixels in the 
shadow mask (Schlapfer et al., 2009). 
The method suggested here is mapping shadow areas in 
HRS images of AISA-Dual sensor. An interpretation of 
boresight band is identifying core shadow areas with highly 
negative values and evaluating 'darkening' for each pixel in the 
classification shadow map. This technique provides an external 
shadow map for de-shadowing algorithm of ATCOR-4, allows 
it to skip six steps of shadow mask identification. The proposed 
a fully automatic method was successfully tested on six scenes 
covering different landscapes. The advantage of the presented 
method is that it does not need a human operator, and it is fast 
processing algorithm exclusively relying on the boresight ratio 
calculated band. 
104