2. THE WORKING PRINCIPLE OF THE DPS 
The working principle of the DPS was at first published on the ISP- 
Symposium, Com.III, Helsinki 1982 (Hofmann, Ebner, Nave, 1982). The 
development of the DPS has been supported by the "Bundesamt für Wehr- 
technik und Beschaffung" and the main results of a referring study are 
published in "Bildmessung und Luftbildwesen" in May 1986. 
The DPS-principle will be briefly repeated. It consists of three main 
steps: data acquisition, correlation process and model compilation. 
2.1 Data acquisition 
The imaging process is performed by a three line push-broom scanner. In 
the focal plane of an objective three linear semiconductor sensor 
arrays A, B, C are arranged perpendicular to the direction of flight. 
These sensors scan the terrain according the push-broom principle 
simultaneously, thereby producing three overlapping image strips Ags 
B s C. of the terrain with different perspectives (Fig. 1). The 
strip coordinates are the line numbers N and the pixel numbers m. 
Fig. 2 demonstrates the overlapping image strips in comparison with the 
overlapping images of conventional aerial cameras. 
ABC 
XA, YA NI, 
X, Y, Z, Len A Jn. If ye 
AO 0b pant OT CS + "1 _-@--+ Flight path 
N fOr 
2) jx / Ne 
HA DOR | A 
x »l« 
+ . 
Update point Pj Lens den nes b------=--—- lé ---------- b----- ---» 
| 7 
  
DEM point P, (X,, Yi, Z) Orthoprojection 
Figure 1. Data acquisition principle of the DPS 
This imaging process can be explained as the travelling of imaged 
terrain points across the line sensors A, B, C in the pixels Ma» Mg» Me 
at the scan cycles Ny, Ng, Ne (Fig...3). 
- 346 -