\Z are
rial photo.
photo using
hen ps and
help of the
rectificated
shift vector
ctor S is to
— AZ)
Zz
— AZ ;(3)
be equal to
limits. For
the scale of
or period T
.1 and R2
] in basic
has to be
on the L1
ained during
value of the
ing for the
que is the
in the plane
differs from
escribed by
(4)
| photogram.
is less than
exceed 1096
e calculated
the help of
spatial coordinates of the points on the glacier
surface.
It is necessary to note once more common positive
quality of the pseudo-parallax method: to identify
the points on the stereopair photograms needs no
their accordance to definite contours. The point
positioning, both movable and stable, is defined by
its coordinates on the L1 photogram, and
comparison with positioning on the R2 photogram is
done by stereoscopic sighting on stereomodel L1-
R2 surface. When the glacier points are moveable
this surface is pseudo-surface.
Conducted works and used equipment
Photos of Medvezhy Glacier (Fig.2) were made
from repeated aerial surveying along the glacier
lengthwise axis with time intervals from several
days to several weeks. When repeated aerial
surveyings were made, definite requirements as
regards for aerial surveying strip direction,
photography height and time were set. Said
parameters have to be as similar as is possible for
different aerial surveyings. Used aerial camera has
a 200 mm focal length, 180x180 mm frame
dimension and a 1:20000 average photography
scale.
Fig.2. Fragment of the aerial photo of the main
icefall of the Glacier "Medvezhy" in the scale of
1:14000. The photo central area is shown.
301
To digitize data by Desk Jet 3c Scanner with 600
dots/inch resolution contact photoprints from
negatives were used. Because of significant
deformation of the photoprints it is necessary to
carry out detection and allowing for deformation
during preliminary processing and to minimize its
effect on the results. Two methods of deformation
detection and allowing for were tested: with the
help of fiducial marks on the photos and with the
help of the crosses on the flattening glass of the
aerial camera.
The first method is based upon off-line conversion
of the photogram points. Coefficients of conversion
are calculated with the help of fiducial marks
measurements. This method is often recommended
in literature when scanned imagery is measured
(Fryer, 1993). The second method is based on the
crosses measurement and corrections calculation
with the use of the 2nd degree polynom.
Practice has shown that the first method lowers
deformation to values of 2-3 pixels, the second one
- to 1 pixel. This magnitude (0.04 mm in our case)
may be considered as real accuracy of contact print
measurements. Experimental estimation of obtained
vectors accuracy has confirmed that fact. Such
accuracy is able to solve different tasks taking place
in application of measurements of aerial
photograms. Therefore, it may be stated that
scanning with a resolution more than 600 dots/inch
is not reasonable because improved resolution
could not be realized due to effect of residual errors
of the photograms.
Measurements were carried out on IBM PC 486
equipped with stereoscopic spectacles on liquid
crystals with 100 cycles per second working as a
shutter. Design of the spectacles and their software
for synchronized working with PC belongs to IBIK
firm. Such a system provides for maximal use of
the display area and has been tested for PC
operations. As an example, publication (Nobuhico,
1992), may be mentioned. All stereoscopic
measurements are carried out according to
stereocomparator technique.
Software and forms of data representation
Software which. has been developed in the
Laboratory of Aerospace Methods enables us to
provide stereoscopic measurements allowing for the
photogram deformations, in order to reconstruct
models using simultaneous and different time
photograms, to calculate coordinates and shift of
the movable point, to represent velocities field
cartographically by vectors (Fig. 3) and by the newly
developed graphical method, so called "strezhinel"
lines. These lines (créated from Russian word
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996