120
1) Obtaining the coordinates.
2) Photo triangular spreading net of ground control
points.
3) Aerial photos external orientation.
4) DTM production.
5) Aerial photos projection on DTM, contours
identification.
From mathematical point of view, stages 2 through 4
are an integral process which essence consists of
establishing correlation and calculation parallax in
stereopairs, only may be done in sequentially stereopair
by stereopair, that inevitably accumulates
geopositioning error as the distance increases from the
true, geodetic specified ground control points.
Plus, nothing is principally changed in terms of resulting
accuracy, with usage of automatic correlation methods
for mutual photos orientation, instead of manual, as well
as DTM production with digital photogrammetry
systems instead of relief depicting via visual stereo
measurements.
Another issue is utilizing GPS receivers in aerial
surveys for registration of principal point spatial
coordinates. Availability of such data, though
significantly improving the situation (due to reduction of
stereomodel mobility during spatial orientation),
nevertheless can not be deemed sufficient for full
automation of stereotopography process.
2) The second category of problems consists of variety
of well known particular limitations of
stereophotogrammetry method. These are studied in
detail therefore it is no sense in thorough description
here. Let us just mention the following.
The problems of such nature are quite diverse in their
characters, but they all are in general tied with issue of
points correlation on stereopair. In certain cases this
leads to complete inapplicability of the method, for
instance in snowed or sanded landscapes with a full
absence of visual texture. In other cases this problem
puts the results' quality in dependence on the number
of factors like average forest elevation and density
when surveying forestry, or buildings shape when
mapping city landscapes.
Limitation of stereophotogrammetry method emerges
mostly in the most practically meaning applications
connected with survey of complex and full of objects
scenes. Particularly due to this reason, large-scale
mapping of city landscapes with significant share of
multilevel buildings can not be done by exclusively
aerial survey methods, thus forcing massive
involvement for this goal carrying out of on-ground
topographic survey, extremely expensive in city
conditions. Besides, there are season limitations
restricting aerial surveys in presence of significant snow
cover or vegetation with leaves. For most part of the
Russian Federation such limitations only leave 1.5-2
months a year for aerial survey.
Practically, such problems often lead to the serious
deformation of technology that causes doubts about
results correctness. Thus, production of DTM of a big
city area considered as compulsory within
stereotopography method, is deemed such a labor
consuming and expensive task affecting the overall cost
of project that a 'compromise' is offered to use a relief
model taken from existing topographic map of
appropriate scale. Given the extremely low metrologic
quality of existing topography basis in Russia, it is only
left to guess what consequences in future would be
caused by such decisions when doing, for example, a
cadastre system to regulate real estate relations.
2. METHODIC PROBLEMS OF JOINT APPLICATION
OF LASER LOCATOR AND PHOTO DATA
Advantages of laser locator methods in DTM production
are commonly recognized.
However, this does not limit the meaning of laser
location in topography. Consistent development of idea
of combination laser locator and digital photo aerial
survey data makes it reasonable to hope on almost fully
automatic technology of aerial survey data processing
for topographic material production.
2.1. Digital Terrain Model
DTM itself obtained with laser location method, has a
number of obvious advantages comparing with classical
stereophotogrammetric DTM:
♦ Traditionally the accuracy of absolute georeferencing
on WGS-84 spheroid is considered as a main
advantage.
Laser scanners (locators) manufacturers usually
declare an accuracy of 15-30 cm on geodetic elevation.
This value is generally determined by achievable
accuracy of range signal measurement with appropriate
electronic unit. Guaranteed planimetric accuracy is
specified in proportion to survey altitude, and normally
ranges from 1/1000 to 1/2000 of flight altitude, that
refers to value of 25-50 cm for typical altitude of 500 m.
It is clear that above values satisfy accuracy
requirements for the most large-scale topographic
maps. It should be mentioned, however, that these
evaluations are not unconditioned, and only achievable
in the most favorable survey conditions.
Such accuracy of georeferencing can not be achieved
with any other remote sensing equipment. This
determines a major difficulty in attempt of experimental
verification - practically the only way of such verification
is low-productive differential GPS on-ground
measurements.
