with a distance of 200 m from a road to see all the
details (e.g.bridges, tubes, kilometre posts
etc.). These types were interpreted just after flight. By
the results in a working copybook and on images
some marks were written. This information was used
for correction of map manuscripts.
Interpretation flight with the second type of the
technology for correcting a map of scale 1:10 000
had a number of special things,got from an absence
of materials of new airsurvey from MDP.Observation
for keeping duty map manuscript by correcting some
details in a copy during flight and landings was
realized.Besides an interpretator had results of
videofilm and map manuscript corrected by
materials of cartographic meaning.Altitude of the
flight was not higher 100m with min speed of 60 km
/h. There were much more landings and virages(by 2
times) comparing with the first type of the
technology.In interpretation works map manuscript
was compared with condition of area some required
detaile and characteristics were done.
Landings of MDP were realized in areas of objects
with difficult situation for interpretation with small
elements that were not recognized in flight. The data
obtained was processed immediately after
observations, while the terrain details are fresh in
the memory of an image interpreter. Each time MDP
flights were followed by field data processing. Field
observations of all the areas flew over took place to
check air visual interpretation data that resulted in
the recognition of about 2096 of objects, thus proving
the interpretation results by photokeys for populated
sites of the second and the third areas. (See Table
2). Field surveying was followed by in-door revision
of field maps and by their preparation for copying
and printing.
2.6 Timely revision of plans and
maps and results evalution.
The 1:500 scaled plans were fully prepared for
revision (control points taken from the plans of 1:500
tacheometric survey, 1993, were repinned). Each
stereo-pair was supplied by, at least, four control
points, by "nests" of contour points of the old plan or
by photogrammetric bridging points (3-4) positioned
at the corners of the area surveyed. Marking device
MP-1 was used for point marking. Control points and
photogrammetric bridging points taken from 1:4000
helicopter air survey documents,1992, were
transfered to air photographs obtained from new air
surveying. As for contour points they were
transfered by repinning from 1:2000 plans made on
the basis of these documents.
Elements of innner and reciprocal orientation were
determined from the models designed on
Stereoanagraph-2, and then tilt angles of MDP air
324
photographs were defined by those elements.
Determination by 11 stereopairs of 1:6000 scale
resulted in Axmax=2°.6, Aomax=7°-4 AXmax=T°
.O, (witct Aomin=0°.1,A0min=0°-0, AXmin=0°.1),and
by 9 stereo-pairs of 1:3000 scale respectively:1.°5,
7.°3, 4.°3 (0.°1, 0.°3, 0.°4).From observation results
you can see that. there are rather significant
differences of tilt angles and turns of air photos
presented in some cases, that is why an analytical
device was used for image processing. Spatial
network bridgings were carried out by an analytical
method with using SK 18x18 stereocomparator and
"Onega-2"automated recording device. The bridging
results used for revision of 1:2000 scaled plans
showed that residual errors do not exceed 0,15mm
with an admissible value being 0,25 mm. As for
bridging based on 1:3000 aerial photographs
intended for revision of plans of 1:500 scale, it failed
to give positive results (errors reached 0,4 mm),
therefore air-photos of 1:6 000 scale were used for
revision on the basis of 1:500 tacheometric survey
data.
Map revision methods vary depending on customer's
demands, so the kinds of revision also may be
different: a partial revision where high accuracy of
some map elements is required; a simplified revision
where accuracy of the whole map or of some map
elements is not so high; and a full revision where it is
necessary to revise the whole map content with a
required accuracy. Based on the results obtained it
was decided to make timely revision of plans at
scales of 1:500, 1:10 000, and 1:2 000 by three
methods respectively.
The choice of a method applied to map revision on
AFP depended on the amount of corrections to be
made, relief of the terrain, and tilt angles of air-
photos. All the working documents i.e data referred
to the flight strip and needed for the map board
orientation, a catalogue of coordinates, and
stereopairs to be processed have been prepared for
the revision. The catalogue of coordinates for the
second area contained 89 control points, for a
processed stereo-pair of 1:6 000 scale almost fully
covering the plan of 1:2000 scale it contained 15
control points. R.m.s. error (further "error") of
reciprocal orientation did not exceed 5 mkm, errors
of exterior orientation computed by 7 points were
dx-0,08 m, dy-0,10 m, dz-0,15 m. Accuracy
estimation of geometrical model designing on the
basis of 14 points resulted in the following values:
dxz0,17m,dy-0,18m, dz-0,25 m. Errors of the map
board orientation were dx=0,12 mm, dy=0,05 mm,
ds=0,13 mm. For the repeated orientation of the
map board five or six identical points per each
stereo-pair were additionally included into the
catalogue. After the evaluation of geometric
constructions completed a digital manuscript of
changes was formed together with simultaneous
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996
correc
correc
and tl
copy.
perfor
New €
By m
manu:
the le
revise
check
data,
1:200!
have
obtain
fullnes
Perce
interpr
with t
demar
numbe
and tl
Issue,
3.C
Instruc
comp
results
objecti
