International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 
  
— 
Fig.3: Additional example of double mapping effects 
1.3 Whatis "True Ortho"? 
Here, the term "True Ortho" means a processing technique to 
compensate for double mapping effects caused by hidden areas. 
It is possible to fill the hidden areas by data from overlapping 
aerial photo images or to mark them by a specified solid colour. 
1.4 Traditional "True Otho" 
In recent years several orthographic rectification schemes 
compensating for double mapping effects were proposed for 
generating large-scale true orthophotos (see, for example, 
F. Amhar 1998 or J.-Y.Rau 2000). These schemes use image- 
based hidden area detection algorithms (modifications of Z- 
buffer). 
Before the orthorectification process begins, a Z-buffer is 
generated. The Z-buffer is a matrix having the same resolution 
as original aerial image. A distance from projection centre to 
elevation surface and the surface primitive identification code 
are stored for each pixel. Each polygon constituting the 
elevation model is projected into the original image plane. The 
projected polygon is then rasterized. That is, for each pixel 
covered by it, a distance from projection centre to the 
unprojected polygon is calculated. If the distance is less than 
the distance value already stored in Z buffer, then the distance 
and identification code are updated. In result, for each pixel of 
the original image, an identification code of the polygon visible 
through it is found. Then the identification code matrix is used 
during the orthorectification process to determine whether the 
particular orthoimage pixel is really visible from the original 
aerial image. 
2. POLYGON-BASED SOLUTION 
In this paper, usage of a polygon-based hidden area detection 
algorithm is proposed. It avoids the generation of large 
auxiliary distance/identification code matrices. 
  
530 
2.1 True Ortho Generation Workflow 
I. Generate conventional orthophoto image 
2. Detect hidden areas (parts of ortho image footprint 
not available on original aerial image). 
3. Use available mosaicking software to fill the areas on 
orthophoto image by a specified color or by raster data 
available from overlapping images. 
2.2 Hidden Area Detection Algorithm 
1. Combine various available elevation data into a 
common polygonal surface. 
2. Project polygons onto source image plane. 
3. Find complete overlay face arrangement. Intersecting 
the projected polygons results in an image place 
subdivision into faces. List of overlapping polygons is 
calculated for each face. 
4. Find the visible polygon for each face. It is easy 
because within a face polygons can be unambiguously 
ordered by distance to projection centre. 
5. Project each face back to polygons visible through. 
The set of the back-projected faces defines part of 
elevation model visible on the original aerial image. 
6. To get hidden areas, project visible parts of elevation 
model onto orthoimage plane and subtract them from the 
orthoimage footprint. 
It is important that no raster data are used. Camera orientation 
parameters and elevation model only are needed for the 
algorithm. 
Potentially the algorithm can be very time consuming because 
intersection of projected polygons can result in the number of 
faces about the squared number of polygons. However, in 
practice the number of faces is not significantly larger than the 
number of polygons, because projection centre (aircraft) is far 
above the urban landscape. So the algorithm takes just a 
fraction of time needed for the alternative Z buffer generation. 
Fig. 4: Orthophoto with hidden areas marked blue 
  
  
  
Fi 
Q 
3.