FUTURE ORIENTED SCENARIOS
FOR
CLOSE-RANGE PHOTOGRAMMETRIC INSTRUMENTATION
Henrik
Resear
S
r
Scientist
cal Resear
Centre of Finl
2150 ESPOO
Finl
ssions I
1
ABSTRACT
Re
closeEach scenario
p
i
scenarios is
rammet
2
INTRODUCTION
seve
te
s
instrumentation.
sts
in usi
ree-
stion is:
re
internal creat
in
r
r our instruments?
rammetrists, our
3
In
FIELD ANOMALY RELAXATION METHOD
techni
0
The
I
FAR is
scri
1977) .
The method
introduced by
Institute
ori
to
recasti
Sli
it
s
r
res research
fields (e
is seemi
lpful in b
ng
of poss
Limitations i
less p
or
socie
later also
1987).
It
j
ns
normally 6 to 10 critical
are
ect
i
s
These
lected to
then arrayed due to possi e conditions whi
are
build up a morphological table. The self-consistent sections as
tion
tempora
snap-shots are
fi
ning a si
together
from
component area.
se
shots are st
into scenarios that seem plausible
the forecas period
and the main stages are dated.
For the pu
recasti
the scenarios for close-range
photogramme ric instrumentation the
rocedure was here
sligtly modified. The number of criti
components was limited
to only f
,each
ti
with
r elements.
re
tempora
the
but also
consis
of the
signifi
conditions
4
MORPHOLOGICAL TABLE
In the following section the morphological table (Table 1) is
presented as a list of the technical, scientific and economi
components which characterize the different stages in photorammetric instrumentation act
The conditions within
criti
are
explained by practi
es.
I.
Image Acquisition vs. Input Image
A.
on
Momentary / Analog. The images are exposed momenta
photographic materi s
the anal
photographs are
used directly for visual interpretat on in
stereoplotters.
B.
Momentary / Digital. The momentary images are digitized
and the interpretation is mainly performed by computer
means.
comparators and analytical plotters equipped
with image digitizing units and also the satellite
imaging systems are included into this category.
Table 1.
CRITICAL
COMPONENTS
IMAGE
ACQUISITION
VS
INPUT
IMAGE
morphological table.
CONDITION
CONDITION
CONDITION
CONDITION
A
B
C
D
MOMENTARY
MOMENTARY
DYNAMIC
DYNAMIC
/
/
/
/
ANALOG
DIGITAL
DIGITALLY
SCANNED
MECH.
SCANNED
ANALOG
IMAGING
ANALOG
IMAGING
DIGITAL
IMAGING
DIGITAL
IMAGING
TRENDS IN
IMAGING
VS
PROJECTIVE
PROCESSING
ANALOG
PROCESSING
DIGITAL
PROCESSING
OFF-LINE
PROCESSING
ON-LINE
PROCESSING
PRIMARY
APPLICATION
TOPOGRAPHIC
MAPPING
SATELLITE
MAPPING
SATELLITE
TOPOGRAPHY
INDUSTRIAL
PRODUCTION
PERSPECTIVE
MATHEMATICS
ANALOG
MATHEMATICS
DIGITAL
MATHEMATICS
3D-IMAGE
MATHEMATICS
GOOD
GOOD
POOR
POOR
/
/
/
RESEARCH
ACTIVITY
ECONOMY
VS
TECHNOLOGY
PROGRESS
II.
/
FAST
/
SLOW
/
SLOW
/
/
FAST
c.
Dynamic / Di~itally Scanned. The dynamic images are
gathered havlng a continuous access by digital means to
the entire scenery covered by the image acquisition
devices. The ordinary, firmly mounted full frame video
cameras where the frames are digitally scanned envisage
best this category.
D.
Dynamic / Mechanically Scanned.
digitized scenery
is dynamic but the scanning of the scenes is performed
mechanically like in servoed theodolites or in cameras
with scanning imager chips.
Imaging vs. Projective Processing
A.
Analog Imaging / Analog Processing. The analog stereo
images are processed by analog means as in the analog
plotters.
V ... 338
B.
Analog Imaging / Digital Processing. The analog images
are processed by dlgltal means like in analytical
plotters. Also the analog and digital theodolites are
included here.
C.
Di ital Ima in / Of
Processin. The digital
images are processed part yon-line ut the object space
coordinates are derived off-line. Already realized.
D.
Digital Imaging / On-Line processing. The projective
equations are processed continuously on-line accordi
to the image frame frequency. Not yet realized.
III.
IV.
Primary Application
A.
Topographic Mapping. The ordinary and main task for
land surveyors as topographers and rna
performed
plotting aerial stereo images.
B.
Satellite Mapping. Primarily used for small-scale
mapping by means of multi-spectral images and digital
image interpretation. Commonly known as remote sensing
applications.
c.
Satellite Topography. The present stage of digital
stereo plotting by means of high-resolution satellite
images.
D.
Industrial Production. The wide range of different kind
of engineering survey applications. When applied for
quality control, also closely confronted to the modern
manufacturing automation.
The Focus of Research Activity
A.
Perspective Mathematics. The research for complete
understanding of the physical behaviour of projective or
even perspective image formation. Actual for surveyors
already since more than a hundred years, but becoming
again popular due to the complicated machine vision
problems.
B.
Analog Mathematics. The construction of analog plotters
once already solved by mechanical or optical-mechanical
means. The future is unknown.
c.
Digital Mathematics. Realization of projective
transformation by digital means like in analytical
plotters. The extension of the mathematical model to
most comprehensive conformance with the physical model.
D.
3D-Image Mathematics. Three-dimensional image
processing in order to build up space images of objects
and the entire scenery_ A critical and thus popular
problem for dynamic machine vision applications.
V-339
v.
Economi
rcumstances vs
Technological Progress
relates to
c cycles accordi
to
ew as it has been pre
Kone
(1985).
c devel
occurs in histor
es of
rs the last minimum
ng inc
around 1950. The
economic growth is always initiated by new technologies
so
in cycles of about 50
rs. After 30
rs of economic
rowth there
lows a crisis. Duri
s crisis new
chnologies are developed, which a
20 years of
recession generate a new rowth.
to
s the
ial
rters of
nni
950 run
A.
still
Gene
B.
C.
I
Recession begun around the mid of
Poor / Slow
seventies. The
rter is for late implementation of
known te
es whereas we still can't
ly see
the new i
recession during nineties be re a
ns a
2000.
proper time to
act
new inventions in
new implementation swi
D.
5
with
es.
SCENARIOS
rammet
5.1
in
As an int
tion to the
re scenarios of
osephotogrammetric instrumentation a scenario for photogrammetry in
general is first presented. This consists of four successive
periods related to the
f rent quarters of the ongoing
economic cycle. Thus three of them are projected back onto
press of photogrammetric instrumentation
nce 1950
four
one is a forecast of
reo
near
The combination of the uppercase letters within each section
indicates the chosen conditions of the five cri cal components
as stated in
mo
table. The conditions are in
o
r according to
there is one choice
r
component respective
The scenario for photogrammetic instrumentation in gene
i.
is:
Around 1950 - 1962: AAABA. The quarter of final
implementation of analog aerial photogrammetry. The main
applications are within topographic plotting over la
areas by aerial analog plotters. The research on
og
processing of stereomodels including analog triangulation
is popular although the pure analytical models are
a)
b)
ramme r c
ose-
ns rumentat
majori
will be on
i
tri
manufacturi
e
is still more or less recessive
of the scenario t
it becomes pro
rous
The te
press rema ns fas all
scenarios
r
ose-
rammetr
trumentat
rato s
ii
BDDCDe
i
more
ri
tems
processi
are
HIGH-ACCURACY
PHOTOGRAMMETRY
Figure 1.
DIGITAL
STEREOPLOTTING
DIGITAL
PHOTO-GEODESY
PERSPECTIVE
PHOTOGRAMMETRY
The scenarios for close-range photogrammetric
instrumentation.
5.2.2
i.
ii.
iii.
5.2.3
Digital Ste
aCCCD. The recent
site satellite
s
s has
rowing need of digital
stereopl
rs.
i
has been known
already for several
rs
some p
shave
under development (e.g. Sarjakoski, 1981;
rtz, 1986).
The analytical plotters are first used for industrial
measurements where the momentary
og
s are
proces
partly
means (Schewe,
88).
BDCDD.
more effici
the systems.
satellite topog
ins
lations al
prima
ication s
indus ria
ects in
manufacturi
process
wide p
ion of
photogrammetric
tems.
ementi
s into
BDDDD. The new st
to
the use of digital
plotters in idustrial appli
ions relies on the
replacement of analog image memories by dynamic digital
images. Incidentally, the scenario of the digital
steroplotting
the scenario
high-accurate
fore
last
DDDDD.
re
ications of
Digital Photo-Geode
i.
DCDCD. First servoed di tal theodolite systems have
been already equipped wi
insi
mounted imager chips
(e.g. Gottwald et al., 1987a
1987b). These
high-accurate systems
developed for
real-time
neering su
mentioned
comparato
photogramme ric
The first
installations exist within the
i
t
ii.
DDDCD.
systems are further
I
by
more efficient hardware and software
r image
processing. Problems caused by the
instabili
of the instrument and object set-ups are
sequential check procedures. The rapid growth 0
number of istalled systems,
predicted by the market
analysts after the first successful reference
installations, is still limited by efficiently propagated
comparator-based photogrammetric systems.
iii.
rst
thus practicable
lations
r
rate applications are among
the
c monitori
of
ree-dimensional object
deformations and
racking control of remotly moving
objects.
new st
is
so leaning upon a growing
resea
of perspect
p
ection mathematics for the
pu
of dynami
ly process the scanning images.
DDDAD.
to
Because
rojection
first after
1
2000
5 .. 2 .. 4
i.
Per
ct
rammet
The real-t
pho
rammetry has
ready become
ac
area
basic research within the society and
rimarily concent
to the geometric processing of
ine
s and to the electronics of image forming
s (e.
1987)
first stand-alone
rammetric
tems have
already
s (
et. 0' 1987).
tric
CCDCD.
e
ii.
iii.
I
e
limitation is the seemingly pure
id state imagers.. This leads first
to la
r
s
es
ti-station imagery and
shortly later
the implantation of advanced imager
technology with higher resolution in the system.
Simultaneous
are also enforced by more
efficient
ine
capacity. The nomi
rates practically
are al
50,000 or even better..
first
tems are ins
led in automated factories,
procedured in accordance
processes .
CDDeD.
CDDDD.
on three-dimensional image
without external
ng amount of installed
tems wi
n the manufacturing industry
to new principles in designing new
r roduction processes already before
cameras ac i
as permanent gages and
s
read
lover the assembly
z
in 0
r
fulfil
are dominant to the
rent
industrial
6
CONCLUSION
The scenarios and the plausibility judgements presented here are
- and as stated already before - merely subjective and don't
tend to be complete in describing the different snap-shots.
Thus they may be treated more as rare examples and alternative
results when applying the modified FAR-method than as personal
statements. The FAR-method proved it's efficiency in opening
ones eyes and it will potentially force everyone to generate his
own futures analyses. The FAR-process also hopefully helps us
to
our creativi
into the areas which are relevant to us
as photogrammetrists.
, according to but also in spite of the above scenarios
re exists an evident trend in close-range photogrammetric
trumentation. The scenarios converge more or less rapi
in
rea
constructions where
dynamic images are direct
scanned in order to produce a continuous flow of threedimensi
cont
in
ion of industrially manufactured
ects. Only
st
es differ from each other depending
on
presently existing
chnological origins respectively.
Conclus
ly the common need
r all these treatments within the
photogrammetric society will be the active reasearch on
geometric processing of three-dimensional digital imagery and
this should further be supported by new developments in dynamic
perspective mathematics. We are coming closer to the optical
photogrammetry or better to say - closer to the perspective
photo!!:grammetry.
REFERENCES
1.
rtz, J., 1986. Digitale Bildverarbeitung in
r
Nahphotogrammetrie - Neue Moglichkeiten und Aufgaben,
Bildmessung und Luftbildwesen, Vol. 54, No.2, pp. 34 - 45.
2.
Brown, D. C., 1987. AutoSet, an Automated Monocomparator
Optimized for Industrial Photogrammetry, a paper presented
at the International Conference and Workshop on Analytical
Instrumentation, Phoenix, Arizona, November 2-6, 1987.
3.
Fraser, C. S., 1988. Periodic Inspection of Industrial
Tooling by Photogrammetry, Photogrammetric Engineering ___
Remote Sensing, Vol. 54, No.2, pp. 211 - 216.
4.
d, R., and Berner, W., 1987a. A New Angle on 3-D
Coordinate Determination, Sensor Review, 7(3), pp. 143 146.
5.
Gottwald, R., and Berner, W., 1987b. Electronic Theodolites
- Sensor-Systems for Real-Time-Photogrammetry ?, Fast
Processing of Photogrammetric Data, published by the ETH
Zlirich, pp.343 - 349.
--
1