Sandau R. Digital Airborne Camera: Introduction and Technology
Подождите немного. Документ загружается.
7.3 UltraCam, Digital Large Format Aerial Frame Camera System 325
7.3.5.2 Aerial Triangulation by Check Points
Another widely accepted method to verify and assess the geometric performance of
mapping cameras is the use of check points. We use the results from six individual
flight missions and six individual cameras to analyze the geometric performance
of these cameras. With an average of 199 check point measurements, deviations
of ±38, ±46 and ±56 mm in X, Y and Z respectively were observed at the GSD
of 10 cm. The vertical accuracy of this dataset corresponds to 0.04‰ of the flying
height (cf. Fig. 7.3-15).
Fig. 7.3-15 Residuals of check points after the bundle adjustment. Results from six individ-
ual cameras and flight missions at 10 cm GSD were merged and 199 measurements from 35
checkpoints were analyzed
7.3.5.3 Geometric Improvement by Auto-Calibration
The end-to-end digital workflow of the UltraCam processing chain offers several
auto-calibration options. One specific function is already implemented in the post-
processing software to reduce the effects of thermal changes during a flight mission
(cf. Fig. 7.3-16). In addition, we have observed positive effects when traditional
radial symmetric distortion parameters are introduced into the bundle solution. Such
parameters are well known in the community, almost every bundle solution offers
them and almost any digital workstation is able to accept such parameters. The
magnitude of these parameters is small (<0.5 pixel), but their effects are systematic
and can therefore not be ignored. In addition to the widely known radial symmetric
distortion parameters we have investigated the positive effect of camera-specific
parameters. Figure 7.3-17 illustrates how the additional parameters are being used.
These parameters are introduced in order to compensate for any asymmetric changes
of the geometry of the camera. Again, the magnitude of these parameters is expected
to be rather small but will play a role when high accuracy is needed. The remaining
326 7 Examples of Large-Scale Digital Airborne Cameras
Fig. 7.3-16 Scheme of the UltraCam stitching concept, which is implemented in the image post-
processing in order to transform the image information from the four camera heads into one single
frame. The two scales shown on the right are computed during this step and compensate for any
dimensional changes of the camera body
Fig. 7.3-17 Image residuals after the aerial triangulation without additional self-calibration
parameters, showing a maximum of 1.67 μm(left). Applying radial symmetric distortion param-
eters (middle) and camera-specific parameters (right) reduces the remaining image distortions to
the sub-micrometre level
residuals in the images have a maximum at ±1.67 μm when no additional self-
calibration parameters are applied. When radial symmetric distortion parameters are
applied, this value decreases to ±1.5 μm and, finally, when additional asymmetric
parameters are also included, to a very accurate value ±0.73 μm.
With the application of automatic self-calibration, users are assured that system-
atic image errors are stable and do not change during the flight mission for a specific
image block. This stability is a clear advantage of the digital camera over the film
camera. The achievement of significant increases in accuracy by means of additional
parameters testifies to the stability of the errors with time.
7.3 UltraCam, Digital Large Format Aerial Frame Camera System 327
7.3.6 Radiometric Quality and Multispectral Capability
The UltraCamX exploits the radiometric quality of the high performance CCD
sensor FTF5033 manufactured by DALSA and preserves this quality by means
of in-house proprietary electronics circuits. No less than 13 bits of radiometric
information can be extracted via the 14-bit analogue/digital converter. Such broad
bandwidth allows the resolution of dark and bright areas in one and the same scene,
for example from a city area on a bright sunny day with dark shadows in the streets
and almost white roofs or other bright objects. The performance in dark image
regions shows the full potential of the sensor and its sensitivity. Only ±6DN@
16 bit (=±0.4 DN @ 12 bit) of noise could be detected in shadows.
Figure 7.3-18 shows frame 1,090 from a flight mission over the city of Graz,
Austria on 27 March 2007, a clear sunny day. At a flying height of 900 m above
ground level a ground sampling distance (GSD) of 6.5 cm was achieved. Two
sub-areas of the panchromatic camera image containing very bright objects (umbrel-
las and welded roofs) as well as dark shadows were analyzed by computing the
histogram. Levels of intensity from 350 DN to 7,800 DN @ 16 bit could be
detected, yet image areas were not saturated. Such huge dynamic range of 7,450
DN corresponds to almost 77 dB or 12.9 bits.
Fig. 7.3-18 Frame 1,090 of an UltraCamX on flight mission over the city of Graz. Two details of
the image show a radiometric bandwidth of almost 13 bits. The darkest pixels are at a level of 350
DN and the brightest pixels are at a level of 7,800 DN in a 16-bit representation
The multispectral channels of the UltraCamX offer true colour (red, green
and blue) as well as near infrared bands and therefore support multispectral
classification, which benefits from a large radiometric range.
328 7 Examples of Large-Scale Digital Airborne Cameras
7.3.7 The Potential of Digital Frame Cameras
The driving force behind innovation in aerial photogrammetry for about 100 years
was to minimize the number of photographs for a given project. The reason was
simply cost, which was directly proportional to the number of images to be flown,
which determined the film to be purchased, processed in a photo-laboratory, per-
haps scanned into a digital format, triangulated and finally manually converted
into information by photogrammetric procedures. The costs for aerial triangulation,
DEM generation and orthophoto processing were a linear function of the number
of images in a project. Cutting the number of flight lines in half meant a reduc-
tion in the cost of a project by the factor of four. The rapid acceptance of digital
cameras is the result first of all of an increase in image quality and secondly of the
economic advantages over the continued use of film cameras, simply because there
is no longer a need to purchase film, develop it and scan it into pixel arrays. Higher
image quality produces a better product, or supports a reduction of flight lines by
flying at higher altitudes. This drives aerial photogrammetry away from film-based
workflows and towards fully-digital approaches.
The third reason for acceptance, however, is the most important factor: the indi-
vidual digital image does not produce any variable cost. Therefore the operation
of a digital airbornee camera is feasible with only fixed annual costs, indepen-
dent of the number of images being produced, be this 1,000 or 100,000 images
per year. Only the costs for the fuel and fees in using the aerial camera platform
are variable. The result is an opportunity to automate photogrammetric procedures
that previously could not be efficiently automated. The many past failures to auto-
mate photogrammetric procedures can be attributed to, firstly, the limitations of film
quality in the form of grain noise and a limited dynamic range; secondly, economy
in the use of film, preventing the creation of sufficiently redundant image blocks
properly to support automation. Having each terrain point on only two images
for stereo does not represent sufficient support for robust automation, where per-
haps five images are needed. Of course, producing a larger number of images only
makes sense if subsequent processing efforts for those images are also indepen-
dent of their number. Therefore automation is the prerequisite for a free choice
of the number of images, and this free choice is a prerequisite for successful
automation.
It is obvious that the future of modern photogrammetry lies in a fully auto-
mated workflow. Automated aerial triangulation may use a vast redundancy of
1,000 or more tie points in every image; the creation of a DEM may rely on
80–90% forward overlap and 30–80% sidelap to look in between vertical objects
and match five images instead of a mere two; the orthophoto may be in the
form of a 3D photo-texture placed on top of the 3D object surface, with classi-
cal orthorectification being nothing more than a special case of the more general
“three-dimensional orthophoto”. Finally, 3D vectors for GIS may be derived from
land use classification and from image analysis, which may also support the
classical processes of aerial triangulation, DEM generation and 3D orthophoto
creation.
7.3 UltraCam, Digital Large Format Aerial Frame Camera System 329
Digital large format airborne cameras help photogrammetry to “strike back”, so
to speak, after a period of retreat where airborne laser scanning took over the gen-
eration of point clouds for DEMs and GPS /IMU compromised the role of aerial
triangulation. The lack of any variable cost for a digital image changes the rules of
photogrammetric imaging and mission planning. This leads to automation opportu-
nities that will invigorate photogrammetric innovation, offer alternatives to airborne
laser scanning for the generation of point clouds and re-establish aerial triangulation
as the central tool for precise image block formation for subsequent multi-image
analyses. All this will also lead to opportunities to create new types of data and thus
“content” for the ever-expanding information society.
It will be the “affordability due to automation” that will make the global
3D databases for “locational awareness on the Internet” a practical reality.
Photogrammetry may gain in relevance by moving centre-stage in the world
of high-bandwidth digital geospatial data creation and broadband Internet-based
delivery.
7.3.8 Microsoft Photogrammetry
The acquisition of Vexcel Corporation, Boulder, Colorado and Vexcel Imaging
GmbH, Graz, Austria, by Microsoft Corporation, Redmond, Washington raised
questions, expectations and guesses in the photogrammetric community when it
was announced in May 2006. The acquisition was driven by Microsoft’s “Virtual
Earth” initiative. The target of this program is to model in three dimensions the
humanhabitatoftheworld–afewthousand cities and their neighbourhoods as
well as rural areas between cities. Photogrammetry is the prime technology to do
this job; the UltraCam is the preferred airborne sensor to build an automated 3D
urban building modelling pipeline.
7.3.8.1 Aerial Missions
Aerial photo missions are carried out at a large scale (six-inch GSD) at end-laps of
80% and sidelaps of 60%. This results in a much larger set of aerial images than was
previously typical in aerial photogrammetry. This highly redundant image data set,
however, supports the automatic processes of aerial triangulation, digital elevation
modelling and feature extraction. Redundancy also ensures the robustness of the
process and the automatic detection of blunders, mismatches and outliers. Another
important advantage of the large overlaps is the rigorous reduction of occlusions,
a most helpful side effect when working in densely built up areas of city centres
(Fig. 7.3-19).
7.3.8.2 Best Visuals for Non-Expert Users
Throughout the history of photogrammetry, geospatial data was produced by experts
to be presented to and used by experts. Depending on the mapping scale a specific
set of symbols had to be used to ensure that maps remained readable, details visible
Abbreviations
AC Alternating Current
AWAF Area Weighted Average Frequency
AWAM Area Weighted Average MTF
AWAR Area Weighted Average Resolution
CCD Charge Coupled Device
CTE Charge Transfer Effectivity
DB Dynamic range
DC Direct Current
DEM Digital Elevation Model
DFT Discrete Fourier Transform
DSNU Dark Signal Non-Uniformity
EWAF Equally Weighted Average Frequency
EWAM Equally Weighted Average MTF
EWAR Equally Weighted Average Resolution
FEE Front End Electronics
FET Field Effect Transistor
FFT Fast Fourier Transform
FMC Forward Motion Compensation
FOV Field of View
FPM Focal Plate Unit
GPS Global Positioning System
GSD Ground Sample Distance
IFOV Instantaneous Field of View
IMU Inertial Measurement Unit
INS Inertial Navigation System
IR Infrared
MS Multispectral
MTF Modulation Transfer Function
NIR Near Infrared
OTF Optical Transfer Function
PAN Panchromatisch
POS Position and Orientation System
PRNU Photo Response Non-Uniformity
333
R. Sandau (ed.), Digital Airborne Camera, DOI 10.1007/978-1-4020-8878-0,
C
Springer Science+Business Media B.V. 2010
334 Abbreviations
PSF Point Spread Function
PTF Phase Transfer Function
RGB Red Green Blue
SNR Signal to Noise Ratio
TDI Time Delayed Integration
TFOV Total Field of View
Bibliography
Ackermann, F. (1995): Digitale Photogrammetrie – Ein Paradigma-Sprung. ZPF, 3, 106–115
Ackermann, F., J. Bodechtel, F. Lanzl, D. Meissner, P. Seige, H. Winkenbach and J. Zilger
(1991): MOMS-02/SPACELAB D-2: A high resolution multispectral/stereo scanner for the
2nd German Spacelab mission. SPIE, 1490, 94–101
Albertz, J. (1999): 90 Jahre Deutsche Gesellschaft für Photogrammetrie und Fernerkundung e. V.
PFG, 5/1999, 293–349
Albertz, J. (2001): Einführung in die Fernerkundung – Grundlagen der Interpretation von Luft-
und Satellitenbildern. 2. Aufl., Wissenschaftliche Buchgesellschaft, Darmstadt
Albertz, J. and W. Kreiling (1989): Photogrammetrisches Taschenbuch. 4. Aufl., Wichmann-
Verlag, Karlsruhe
Analog Devices (2002): AD9824-Complete 14-Bit 30 MSPS CCD Signal Processor, Rev.0,
© Analog Devices, Inc., 2002
ATMEL (2001): TH7982AVWB90NB Detail Specification Correlated Double Sampling for Space
Applications, ATMEL Grenoble, 22.11.2001
Bähr, H.-P. and T. Vögtle (Hrsg.) (1998): Digitale Bildverarbeitung – Anwendungen in
Photogrammetrie, Kartographie und Fernerkundung. 3. Aufl., Wichmann Verlag, Heidelberg,
373 S.
Bass, M., E. W. van Stryland, D. R. Williams and W. L. Wolfe (Eds.) (1995): Handbook of Optics.
2nd Ed., McGraw-Hill, New York
Bäumker, M. and F.-J. Heimes (2001): New calibration and computing method for direct georef-
erencing of image and scanner data using the positions and angular data of an hybrid inertial
navigation system. In: Heipke, C., K. Jacobsen and H. Wegmann (Eds.): Integrated Sensor
Orientation – Test Report and Workshop proceedings. 16 S.
Bormann, G. E. (1975): Measurement of Radial Lens Distortion with the Wild Horizontal
Goniometer. ISP Commission I working group study
Brieß, K., W. Bärwald, F. Lura, S. Montenegro, D. Oertel, H. Studemund and G. Schlotzhauer
(2001): The BIRD-Mission is completed for launch with the PSLV-C3 in 2001. In: H. P. Röser,
R. Sandau and A. Valenzuela (Eds.): Small Satellites for Earth Observation III. Wissenschaft
und Technik Verlag, Berlin
Bronstein, I. N. and K. A. Semendjajew (1989): Taschenbuch der Mathematik. 24 Aufl., Verlag
Nauka, Moskau und BSB B. G. Teubner Verlagsgesellschaft, Leipzig
Brown, D. C. (1971): Close-range camera calibration. Photogrammetric Engineering, 37(8),
855–866
Brown, D. C. (1976): The Bundle Adjustment – Progress and Prospects. XIIIth Congress of the
ISP, Commission III, Helsinki
Brown, R. and P. Hwang (1992): Introduction to Random Signals and Applied Kalman Filter ing.
J. Wiley and Sons, New York, 502 S.
Coffey, K., J. Warta and A. Krämer (1999): Rauschverhalten hochauflösender AD-Wandler.
Elektronik-Industrie, 11/1999
335
336 Bibliography
Cramer, M. (1999): Direct geocoding – is aerial triangulation obsolete? In: Fritsch, D. and
R. Spiller (Eds.): Photogrammetric Week ’99. Wichmann Verlag, Heidelberg, S. 59–70
Cramer, M. (2003): Erfahrungen mit der direkten Georeferenzierung. Zeitschrift für
Photogrammetrie, Fernerkundung und Geoinformation PFG, 4/2003, 267–278
Cramer, M. (2004): www.ifp.uni-stuttgart.de/eurosdr/cramer-eurosdr-isprs.pdf
Datel (2003): Applikation zur Messdatenerfassung
Derenyi, E. E. (1970): An exploratory investigation concerning the relative orientation of continu-
ous strip imagery – A thesis submitted in partial fulfilment of the requirements for the degree
of Doctor of Philosophy in the Department of Surveying Engineering of the University of New
Brunswick, March 1970
Dierickx, B. (1999): Electronic image sensors vs. film: beyond state-of-the-art. OEEPE
Workshop on Automation in Digital Photogrammetric Production, 21–24 June 1999, Paris.
http://www.fillfactory.com/htm/technology/pdf/oeepe99.pdf
Dimac (2004): www.dimacsystems.com
Ebner, H., D. Fritsch, W. Gillessen and C. Heipke (1987): Integration von Bildzuordnung und
Objektrekonstruktion Innerhalb der Digitalen Photogrammetrie. BuL, 55, 194–203
Eckardt, A. (2002): Design und Verifikation der ersten digitalen kommerziellen Luftbildkamera
(ADS). Dissertation, TU Berlin
El-Sheimy, N. (2003): Inertial techniques and INS /DGPS integration, lecture notes, ENGO 623
course, Department of Geomatics, The University of Calgary, Calgary, Canada
Engelmann, M. and N. Ahner (2004): CCD -Technik und Bildaufnahme. Vortrag im Rahmen der
Vorlesung Elektronische Bauelemente an der TU Chemnitz
Fairchild (2004): Spezifikation der CCD525. Fairchild, USA
Falkner, E. (1994): Aerial Mapping, Methods and Application. CRC Press Inc., Florida
Finsterwalder, R. and W. Hofmann (1968): Photogrammetrie. 3. Aufl., De Gruyter, Berlin, 455 S.
Foitzik, L. and H. Hinzpeter (1958): Sonnenstrahlung und Lufttrübung. Akademische
Verlagsgesellschaft, Wiesbaden
Fricker, P., R. Sandau and A. S. Walker (2000): Progress in the development of a high performance
airborne digital sensor. Photogrammetric Record, 16 (96), 911–927
Fritsch, D. (1998): Photogrammetrie und GIS – eine Ehe mit Zukunft. Keynote Paper
Intergraph’s 1. Photogrammetrie-Forum 1998
Fritsch, D. and D. Stallmann (2000): Rigorous Photogrammetric Processing of High
Resolution Satellite Imagery. International Archives of ISPRS, XXXIII(B1), 313–321,
Amsterdam
Gelb, A. (1974): Applied Optimal Estimation. MIT Press, Cambridge, Massachusetts, US, 374 S.
GeoMat (2004): Geographische Materialien, 26, 173–182
Gervaix, F. (2002): Aerotriangulation : auch für den ADS-40 Luftbildsensor? PFG, 2, 85–91
Göhring, D. (2002): Digitalkameratechnologien – Eine vergleichende Betrachtung, CCD kontra
CMOS , 12.08.2002. Technische Informatik, Humboldt Universität zu Berlin
Goodman, J. W. (1985): Statistical Optics. J. Wiley and Sons, New York
Gotthard, E. (1975): Zusatzglieder bei der Aerotrangulation. In: BuL 43, 218–221
Graham, R. W., W. S. Warner and R. E. Read (1996): Small Format Aerial Photography. Whittles
Publishing, Scotland
Greening, T., W. Schickler and A. Thorpe (2000): The proper use of directly observed orientation
data: Aerial triangulation is not obsolete. In: Proceedings ASPRS annual conference, on CD-
Rom, 12 S., Washington DC, May 22–26
Grewal, M. and A. Andrews (1993): Kalman Filter ing: Theory and Practice. Prentice Hall, New
Jersey, USA, 382 S.
Gruber, M. and R. Ladstädter (2006): Geometric issues of the digital large format aerial camera
UltraCamD. International Calibration and Orientation Workshop EuroCOW 2006 Proceedings,
25–27 Jänner 2006, Castelldefels, Spanien.
Gruber, M. and U. Manshold (2008): UltraCamX-Large Format Digital Aerial Camera, DGPF
Conference, Oldenburg, April 23-25th, 2008, Presentation
Bibliography 337
Grün, A. (1978): Accuracy, reliability and statistics in close range photogrammetry. Inter-Kongress
Symposium, Com. 5, Stockholm, Schweden
Grün, A. and E. Baltsavias (1986): High Precision Image Matching for Digital Terrain Model
Generation. International Archives of Photogrammetry and Remote Sensing, XXVI, III/1,
284–296, Commission III. Rovaniemi
Gruen, A. and L. Zhang (2003): Sensor modeling for aerial triangulation with Three-Line-Scanner
(TLS) imagery. Photogrammetrie, Fernerkundung, Geoinformation (PFG), 2, 85–98
Heipke, C., K. Jacobsen and H. Wegmann (Eds.) (2001): Integrated sensor orientation – test
report and workshop proceedings, OEEPE official publication No. 43, 297 S., verfügbar unter
http://www.oeepe.org/ (Zugriff August 2004)
Hildebrandt, G. (1996): Fernerkundung und Luftbildmessung. Herbert Wichmann Verlag,
Heidelberg
Hofmann, O. (1982): Dynamische Photogrammetrie. BuL, 3, 3–19
Hofmann, O. (1988): Photogrammetrisches Verfahren für Fluggeräte und Raumflugkörper zur
digitalen Geländedarstellung. Patent DE 294087122, angemeldet 1979, erteilt 1983
Hofmann-Wellenhof, B., H. Lichtenegger and J. Collins (2001): Global Positioning System:
Theory and Practice. Springer-Verlag, Wien, 382 S.
Holst, G. C. (1998a): Sampling, Aliasing , and Data Fidelity. SPIE Optical Engineering Press,
Washington
Holst, G. C. (1998b): Testing and Evaluation of Infrared Imaging Systems. SPIE Optical
Engineering Press, Washington
Holst, G. C. (1996): CCD arrays, Cameras and Displays. SPIE Optical Engineering Press,
Bellingham, Washington
Hydrolab (2004): http://hydrolab.arsusda.gov/rsbasics/noise.php
IAI FZK (2004): Bestimmung der Auflösung von Bilderfassungskomponenten © 2004 – Institut
für Angewandte Informatik; Forschungszentrum Karlsruhe (http://www.iai.fzk.de/projekte/
mikrologistik/bildverarbeitung/module/aufloesung/index.html)
Intergraph (2008): http://img.en25.com/web/Intergraph/DMC.pdf
Iqbal, M. (1983): An Introduction to Solar Radiation . Academic Press, New York
Jacobsen, K. (1982): Programmgesteuerte Auswahl zusätzlicher Parameter. In: BuL 1982,
213–217
Jacobsen, K. (1998): Geometric Potential of IRS-1C PAN-Cameras. ISPRS Com I Symposium,
Bangalore, India
Jacobsen, K. (2002): Calibration aspects in direct georeferencing of frame imagery. ISPRS
Commission I/Pecora 15 conference proceedings, IAPRS (34), Part 1 Com 1, 82–89, Denver,
2002
Jacobsen, K. (2003): System calibration for direct and integrated sensor orientation. In:
Proceedings ISPRS WG I/5 workshop on theory, technology and realities of inertial/ GPS
sensor orientation, on CD-Rom, 6 S., Castelldefels, 2003
Jahn, H. and R. Reulke (1995): Systemtheoretische Grundlagen Optoelektronischer Sensoren.
Akademie Verlag, Berlin
Johnson, J. (1985): Analysis of Image Forming Systems. In: Proceedings of the Image Intensifier
Symposium, 249–273. Warfare Electrical Engineering Department, US Army Engineering
Research and Development Laboratories, Ft. Belvois, VA. Nachdruck dieses Artikels im
Selected Papers on Infrared Design, R. B. Johnson and W. L. Wolfe (eds.): SPIE Proceedings
513, 761–781
Kodak (1994): Solid State Image Sensors_Kodak-Grundlagen_DS00-001, Rev. 0, 08.12.1994
Kodak (1999): KAF- 6303LE 3072 (H) x 2048 (V) Pixel Enhanced Response Full- Frame CCD
Image Sensor, Rev. A, 25.09.1999
Kodak (2001a): ConversionOfLightToElectronicCharge_Kodak_MTD-PS0217; Rev. 1,
29.05.2001
Kodak (2001b): Image Sensor Noise Sources, Application Note Kodak MTD/PS-0233, Rev. 1,
01.08.2001