by an
> first
ed by
nd Z
of the
re the
high
ave a
ISy to
irther
new
2 by
ound
fit is
w Ol
ICP)
Kay,
is d2
stical
ation
-arily
rary
>curs
elow
Martin Smith
A further development of the technique was introduced to increase the number of points in a curve by generating
intermediate points using spline curve fitting.
4 POTENTIAL INFLUENCES ON THE QUALITY OF THE RESULTS
Initially some theoretical, simulated data tests were performed to test the correct operation of the software and to
undertake some basic tests of the algorithms. It was found that the initial difference in position, scale and orientation
was within normally accepted tolerances for photogrammetric processes, for example + 6° of tilt were not a problem.
Practical testing would reveal the effects of:
e the distance between points on the curves;
e the geometry of the curve (shape);
e the quality of the curve definition for example, was the road centre line followed, curb line or somewhere in
between?
Overriding influences on these effects is the problem of ‘dead’ ground on the imagery and, as RTK GPS was used to
produce 3D ground curves the problem of loss of lock on satellites, due to masking of the sky, creating gaps in the
ground curves.
The practical trials were undertaken with three stereo pairs of photographs of different types of terrain, with different
distribution of linear features and with different observing strategies. The process of absolute orientation and exterior
orientation are investigated.
5 RESULTS
The results will be discussed in two sections, the first will deal with the process of absolute orientation where
coordinates from a stereo model will be matched with coordinates derived from GPS measurements. The second section
will deal with results from using curves measured on individual photographs (the case of exterior orientation).
5.1 Absolute Orientation
5.1.1 Nottingham University Campus photography results. The ‘campus’ photography is 1:10 000 scale photography
covering an area of gradual but significant changes in height as can be seen from figure 3.
SA
A
; LX
Campus | M
ll NUS
/ NX
¢ M
Vy >
/ Road
n
/ Ge
pf à 2
3.38... ee Vas
x 10° > ennt a E
327 ur 494 x 10°
Northings i di Eastings
Figure 3. GPS coordinates collected at the Campus site
(note: exaggerated Z scale, all coordinates in metres)
The GPS curve was recorded by walking round the road with a real-time kinematic (RTK) GPS system. RTK GPS can
provide significantly higher accuracy than is required for the ground coordinates for this scale of photography, but this
system does allow easy recording of the curve data by following the kerb line with the antenna. The model coordinates
were observed using the Leica SD2000 analytical plotter.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000. 853