124
.. 'L
NORTH
EAST
> > > \SUISSSXDEMO.ftLX
1 LINE 1 6 ft*
234 AED-28 75.99 4.11 W
235 AED-28 85.64 1,21 6.84
236 AED-28 190.62 -37.66 6.15
237 AED -28 199.53 58.68 5.95 B
238 AED-28 178.56 0.76 29.39
239 AED-28
CUiV 1-lfJUJ UlSi
uz'.hijj; uo:
1010DI
fii-i- UZL GD'i'CJ Zh 1 '1' >S:i I LhlZ
Tel.+41<22>7328810 Fax+41<22>7410450
E-mail: altexPgeneua-1 ink.ch
DIGITAL HAPPING SVSTEMS
5175501.03 i
298118.18
Figure 3.
The final result of data processing in form of photomap. Raster topography map, laser image and orthorectified photo
along with main contours and other semantic data are presented.
1) Full or partial absence of corresponding points on a
pair photo, that might be caused by complex relief,
vegetation, buildings, or other factors. The major
negative result here is malfunction, and sometimes a
failure of 'above algorithms of automatic stereomodel
construction. Thus, even in the current digital
photogrammetry systems these processes require the
most thorough attention of operator to each stereopair.
And this very fact results in such a high cost of
stereophotogrammetric DTM production.
2) Another principal difficulty is that georeferencing
error grows up as a frame distance increases from
primary (geodetic) ground control points, on the above
mentioned of lack of interior control mechanism for such
situation in the stereotopograrphy method. This issue is
less important in area than in corridor survey, and with
available GPS reference of principal point.
Nevertheless, this very reason causes loss of stability
(mostly cross-path) and therefore geopositioning
accuracy reduction in one of the most practically
important cases - corridor survey of long linear objects
(power lines, oil and gas piping etc.) in hard accessible
areas where such geodetic support is extremely
complicated (sometimes just impossible).
So the problem of photos mutual orientation in stereo
pair is the most difficult thing on the way toward the
complete automation of steretopographical method.
Laser locator DTM may play the main role here.
The main procedures of automatic mutual orientation
algorithm are presented below.
1) As a first step the special DTM processing is
implemented with a goal to recognize the fragments
which are “favorable” in terms of possibility of
mentioned above correlation algorithms usage for these
fragments. It means that such fragments must be
smooth enough and be free from vegetation, buildings,
etc.
2) The aim of the second stage is transfer of selected
“favorable” fragments (their mathematical images) to
the coordinate system of each frame taking into account