r all important
otos made by
interest at the
periment and
method of
togrammetric
ot provide all
accident was
| tacheometer
(camera Wild
eras (Mamia
imm, Minolta
| out to serve
considerably
rates then
1eometry, the
rmined from
same points
(tric cameras
nera. Photo
|parator, and
ages BINGO
series, where
camera and
ndent series,
of non-metric
naterial, have
ic survey and
adjustments,
presented in
Jsions were
he achieved
t and photo
r the realised
experiment it is 4-8 cm in XY plane and 1-3 cm in Z-
axis.
2. The accuracy of Z coordinate is up to three times
accuracy of X and Y coordinate.
3. Regardles of theoretical accuracy of image coordinates
measurement which is 1mm, the achieved accuracy is
considerably lower (8-13mm). The reason is found in
the errors of geometrical projection which are not
eliminated, and partially in the errors of point definition.
Table 1 : The results of photogrammetric adjustments
compared with the results of geodetic survey
Camera N c [mm] | m, [mm]
Film Photo! c U &[mm] | m, [mm]
Format no |[mm]| N-U | n [mm] | m, [mm]
MAMIA 204 | 90.163 0.064
black &whit | 5 |0.013| 153 | -0.409 0.076
60x60 mm 51 0.139 0.013
YASHICA 237 | 90.295 0.043
colour 6 |0.013| 162 | -0.226 0.064
60x60 mm 75 0.125 0.013
PRACTICA 202 | 52.481 0.047
black&white] 5 | 0.012} 141 -0.151 0.080
60x60 mm 61 0.548 0.024
MINOLTA 212 | 51.623 0.046
black&white] 5 |0.014| 156 | -1.451 0.037
24x36 mm 56 0.373 0.013
MAMIA 131 | 54.454 0.032
colour 3 |0008| 117 0.012 0.020
24x36 mm 14 -0.012 0.030
At the end of 1993, this experiment was extended
(Mihajlovié, 1994). The task was to take photography of the
same scene accident by using camera RolleiFlex 6006 and
to measure photos and process data with the RolleiMetric
MR2 system. It was also planned to use that system for
photo measurement and data processing for all previously
made photos within the experiment. The purpose of this
extension of the experiment was to answer to the following
questions:
7. Is it possible fo replace comparator measurements
(Which are rather expensive) with tablet digitizer
measurements (which are inexpensive)?
2 7o what an extent is the accuracy improved (in this
particular experiment by using metric camera RolleiFlex
6006 comparing the accuracy achieved by use of non-
metric photos made in 1987 ?
3. Is there à difference between the results of dala
processing carried! out in 1987 and the results of dala
processing using the ROLLEIMETRIC MR2 system ?
In addition to other useful experiences and observations,
the following answers have been obtained:
1. The loss of accuracy deriving from measurements of
photos using tablet digitizer is insignificant. The
explanation is in the fact that comparator
measurements, as the most accurate photogrammetric
measurements, cannot significantly improve the overall
accuracy because very often, the measured points are
not precisely defined. This enables more freedom in
planning and implementation of future projects.
2. With regard to accuracy, the results achieved within the
experiment extension are almost identical to the results
of the experiment that was realised in 1987. The
explanation is to be found in the fact that the increase of
the overall accuracy obtained by use of semi-metric
camera has been neutralised by the lower accuracy of
image coordinates measurements.
3. The results obtained by data processing by use of two
independent software systems are identical. That was
to be expected, considering the same mathematical
model being applied.
In the 1972-1995 period, the experts of the Police
Department in Belgrade were using camera Wild C-120 for
photography of the traffic scene accidents accompanied
with heavy injuries and damages and also for homicide
cases. Restitution of photos was being carried out using
Wild A-40. In the middle of 1995, this equipment was
replaced with the RolleiFlex 3003 camera and the
RolleiMetric MR2 photogrammetric system.
1. TRAFFIC SCENE ACCIDENT ANALYSIS
For court investigation requirements, expertise of the traffic
scene accident that happened in November 1990 was
carried out using available photo documentation (Mihajlovié,
1994). Laboratory for Photogrammetry at the Faculty of Civil
Engineering was appointed to conduct the expertise. The
direct task of the expertise that was to be based on
information contained in documents and available photo
documentation was to determine the following:
- "f traces on the street which are noticeable on photos
from photo oocumentaton belong to the vehicle arven
by the accused,
- Olrection andspeed of the vehicle otiven by the accused
need to be determined using the acquired! data,
- Using available photo documentation investgaie ail the
acts that are important for the time ana spatial analysis of
the trafic accident."
Figure 1: Photo of the traffic scene accident
275
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
