Full text: XVIIIth Congress (Part B2)

m as 
GPS 
1024 
| DPA 
more 
aerial 
rd us- 
jiliza- 
NS is 
Zeiss 
intary 
'order 
4 
‘minal 
od im- 
mage 
ith re- 
lem is 
metric 
seven 
Bing a 
table. 
rected 
1se of 
|. The 
during 
  
ght ex- 
is flight 
ihingen 
iis area 
investi- 
rection) 
| benef- 
Table 2: Parameters of the test flight Mühlacker-Vaihingen 
Area: 4.5 km x 7 km 
Flying height above ground: 2300 m 
Velocity of aircraft: 70 m/s 
Scan frequency: 235 Hz 
Ground pixel size of 
stereo images: 0.30 m 
multispectral images: 0.60 m 
Date of flight: 26.07.1995 
The test flight was performed together with Kirchner and 
Wolf Consult, Hildesheim. An aircraft equipped for carry- 
ing two camera systems simultaneously was used which 
allowed for taking DPA images and aerial images with the 
RMK TOP 15 during the same flight. For our investiga- 
tions this gives us the opportunity to compare the quality of 
photogrammetric restitution using the DPA with the results 
found by the classical photogrammetric techniques using 
aerial images. For having ground truth data about 200 sig- 
nalized points have been spread out in the test region. For 
that purpose white PVC-plates have been used and white 
paintings are drawn on ground with a size of 1 x 1 m?. The 
scene is captured by three strips flown in east-west direc- 
tion and three strips in north-south direction. The neigh- 
bouring strips have a side overlap of 60 percent. 
The scheme for processing the DPA data of this flight exper- 
iment is shown in figure 2. Computations with the 1 Hz GPS 
data are carried out using standard software packages. Po- 
sition and attitude determination from the INS data is solved 
within the DPA system which also serves for HDDT conver- 
sion of the recorded data. The other two boxes in figure 2 
deal with GPS-INS integration and image rectification. For 
these tasks software has been developed. The results of 
this processes are presented in the next section. 
(=) (a) (= ) (os. ) 
aus 
  
  
  
  
  
  
  
  
  
  
  
  
RMK 
Time Stamp DPA ape DPA 
Synchroni- 
Y sation System 
conl 
oordinates 
GPS ad A Time Stamp | HDDT 
Processing Y 
  
  
  
  
Camera GPS Time HDDT 
Position y Conversion 
  
  
  
  
  
Original 
GPS-INS mí an 
Ÿ 
INS Dat 
Integration SD 
  
  
  
  
| > Image 
Camera Orientation Rectification 
  
  
  
(without DTM) 
fr nr nr gon mn ps fs gon ton pean pp ts on nts nt we nem — e— nn 
Rectified Images 
Y y i 
  
  
  
  
  
  
  
  
  
DTMs. 
Digital ; GIS, 
rthoi y 
Photogrammetric ono rap es Thematic 
Stations Mapping Mapping 
  
Figure 2: DPA test flight processing scheme 
The intersecting line between DPA specific data processing 
with the upgrade of raw data and the application depen- 
dent further use of processed data is currently drawn at 
the end of the image rectification process. The rectified 
images can be introduced into Digital Photogrammetric 
143 
Workstations where stereo viewing supports the measure- 
ment processes. Stereo impression is quite bad with the 
directly recorded image data even though a stabilized plat- 
form is used. The rectified image data can be processed 
further with line triangulation modules for 3D recovery or 
with DTM packages, for example, the MOMS algorithms 
(Fritsch et al., 1995) for DTM reconstruction. The geo- 
referenced multispectral and panchromatic images are of 
increasing importance as a source for GIS. Our interest is 
in topographic and thematic mapping with the DPA data 
where a lot of research has to be done in fusing the data 
and the evaluation and interpretation techniques to exploit 
the potential of the DPA sensor. 
4. PHOTOGRAMMETRIC-THEMATIC PROCESSING 
In this section we present experiences with georeferencing 
and image rectification of DPA stereo images and discuss 
first results obtained by multispectral analysis of the spec- 
tral DPA channels. 
4.1 GPS-INS Integration 
The INS is required to measure the high frequency transla- 
tional and rotational motion parameters of the airborne sen- 
sor. Because the optic module is put on a stabilized plat- 
form only angular movements are recorded which are not 
compensated by the platform. A plot of the attitude rates is 
shown in figure 3. Quite easy to see are a few sharp peaks 
with an amplitude of 0.015? at 235 Hz or 3.5? /sec. This high 
changes of the attitudes can only be explained by rotational 
aircraft motion passing over the compensation interval (of 
around £5°) in which the platform is fully operational. 
Roll [degree/ine] 
  
  
  
Pitch [degree/ine] 
  
  
  
  
  
  
  
  
Figure 3: Attitude rates (roll and pitch) of one strip 
For georeferencing of airborne sensors the GPS-INS inte- 
gration is considered to be the most powerful technique. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B2. Vienna 1996 
 
	        
Waiting...

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.