Full text: XIXth congress (Part B7,3)

  
Patikova, Alice 
  
the pictures to enable deleting areas of military protection out of aerial photos. Scanning was done by 1200 dpi. The 
certification of European commission in Ispra with link of kind of use of this data was a good help to get the photos. 
The deliverable of this data had been difficult and needed long time. Data on CD-ROM had to be officially ordered 
from some Slovak company. 
2.3.3.1 Processing of Aerial Photographs 
With the software PhoTopoL operations of the digital photogrammetry are possible. The scanned photos had to be 
prepared by internal and external orientation. 
For internal orientation the co-ordinates of camera fiducial marks are used. The calibration protocols, which were 
delivered together with the photos, have been mostly wrong. Suitable protocols were found in literature in the 
Department of Geodesy of Technical University in Brno. Other co-ordinates have been measured from hard copies or 
by recalculation of image co-ordinates by using resolution of the scanned pictures. 
External orientation was done by use of identical points on photo and map. At each picture between 12 and 20 points 
were picked up and calculation was done by using bundle-adjustment method. The pictures were prepared for 
orthorectification. 
Before the orthorectification of the aerial photos could be done, a digital terrain model had to be created. Because no 
existing model had been available for adequate price or the quality was too bad, the model was created by us through 
vectorisation of contour lines with the software Atlas DMT. The geo-referenced maps out of TopoL-GIS were used in 
the software module NetEditor, which enables placement of points with interactive recalculation of model with its 
contour lines. The created model became very complex and had finally 279.545 points. Other operations out of this 
model, like profile drawing, slope-calculation, volume-validation, 3D-views and other things are possible in AtlasDMT. 
The model can be directly used from TopoL by DDE-connection. 
The new created orthophotos were simplified to 2m pixel-size. The resolution is good enough for identify the changes 
and each photo has less than 20 Mbytes. For comparing, original photo had about 120 Mbytes. 
2.4 Digitalisation and Interpretation of Prepared Data 
Vectorisation and interpretation was done together. At first line-features were mapped out of satellite image and maps. 
Many information from analogue city-plan have been integrated. Topological structure was used and different land-use 
classes have been documented in a database. The vectorisation was done on screen in two windows, in one on basement 
of the satellite-image, in the other one on the maps. Result of mapping was seen interactive in both windows. In the 
other window the original aerial photo was displayed. Some shadows, which made identification and orientation 
difficult, were very often found in the satellite image. Behind of high buildings streets were very difficult to locate. 
Especially in panel house areas, where buildings sometimes guide roads several hundred meters, this effect took place. 
Some of them have dark roofs, which are made of new bitumen-seal. In that case end of building on shadow-side was 
not identifiable. The same effect was found in the areas of water-bodies in the mead-forest area. Shadow and water had 
usually the same grey scale. Some roads are guided from trees on one side. In such cases it is difficult to identify, on 
which side is road and on which is shadow. Big constructions were usually very good for identification on satellite- 
image, with small buildings it was much more difficult. There have been also problems in the central old part of 
Bratislava, where old houses with 3 or more floors stand close together. Other problematic field is the area of cottages 
(or small houses) under forest. When trees are high, the cottages are nearly not visible. Only when there is also use of 
gardens or an agglomeration with higher density, you can identify it in satellite image. Mostly aerial photos and maps 
were used to interpret the land-use of the area. Some official buildings like schools, hospitals and so on, are signed in 
maps with some symbols or cartographic signs (logogram). Special books with description of symbols and logograms 
have been used for map interpretation. Additional to this, field checks have been used for validation of interpretation. 
Photographs have been taken for the use in documentation. Areas were mapped as homogeneous parts with one 
polygon. Setting the label point and filling the database according to MURBANDY legend created the area. Sometimes 
there have been several kinds of use in areas and it was not possible to divide it in the horizontal direction. In that case 
the main area use of all floors or the importance of the area was taken into account. Because of this the residential 
structure seem to be over-represented. 
  
1122 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000.
	        
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.