SURFACE TOPOGRAPHY RECONSTRUCTION OF VENUS FROM THE MAGELLAN MISSION
F. W. Leberl, W. Walcher, R. Kalliany
Institut für Computerunterstützte Geometrie und Graphik
E-Mail: leberl @icg.tu-graz.ac.at
Invited Paper
ISPRS-Commission 4, Working Group 5
KEY WORDS: Extraterrestrial, Satellite, Radargrammetry, Stereo-Radar
ABSTRACT
NASA's Magellan mission to map planet Venus was an unqualified success. Over 95% of the planet’s surface have been
imaged, a large portion of that even more than once. The imaging part of the mission ended in 1992; the satellite itself had its
life terminated in 1994. The images of the planet represent more than 400 Gbytes of pixels, distributed over more than 5000
individual orbits. These images have been entered into a data base of NASA's Planetary Data System PDS. As the major
source of information about Venus, they are being studied now and for many years to come. This paper reports about some
studies relating to managing the images and extracting the detailed topographic surface shape. One exploits the overlaps
among images taken during different phases or cycles of the mission. Major issues in theses studies are the extent to which
terrain features are laid-over at steep look angles off-nadir, the obstruction of analyses by the dissimilarity of images taken
from opposite sides, and the sheer size of the data which causes difficulties in the interaction and use of the images.
ZUSAMMENFASSUNG
NASAs Magellan-Mission zum Planeten Venus wurde ein voller Erfolg. Mehr als 9596 der Oberflüche des Planeten wurden
abgebildet, ein Grofteil sogar mehr als einmal. Die bildgebende Phase der Mission dauerte bis 1992, der Satellit beendete
seine Existenz aber erst 1994. Die Bilddatenmenge umfafit etwas mehr als 400 Gbytes an Pixeln und betrifft mehr als 5.000
Umlaufbahnen. Diese Bilder wurden nun in NASAs Planetary Data System (PDS) eingegeben und stehen für die
Detailanalyse bereit, welche sich noch weit in die Zukunft dieser primären Informationsquelle des Planeten wird widmen
kónnen. Dieses Paper berichtet nun über einige dieser Studien zur Verwaltung dieser Bildmengen und zur Detailanalyse der
Oberflächenformen. Es wird die Überlappung zwischen den Bildern genutzt, welche in den verschiedenen Projektphasen
entstanden. Wesentliche Sorge dieser Arbeiten ist einerseits der hohe Anteil an topographischen Merkmalen, welche wegen
der steilen Beleuchtungsrichtungen in einen sogenannten Layover fallen, weiters die Unáhnlichkeit der überlappenden Bilder
aus unterschiedlichen Beleuchtungsrichtungen, als auch die grofe Datenmenge, welche Schwierigkeiten bei der schnellen
interaktiven Handhabung zur Folge hat.
1 INTRODUCTION
When the radar imaging system on NASA's Magellan Megelan Orbi Elements 10 wot eem s
satellite ceased operations in 1992, it had not only already i Orbit = 1 FBIDR |
assembled an amazing 400 Gbytes of images, covering i
more than 97% of the planet’s surface, but it had also
17000 km
(220000 pixel)
exhausted NASA's ability to keep the data stream funded. ;
Within the almost polar, but highly elliptical Venus-orbit —
of the Magellan-probe, imaging periods were about 37
minutes per orbit. Within this time, a swath of about 20km 9
width and 1700km length was to be acquired. Play-back to
|
|
|
'
Earth took place for more than 100 minutes during the e
more distant section of the orbit, at a transmission rate of Q oes iom
270 Kbits/ d Fi d Daia recording ohase 20 km
Its/second. (see 1gure 1). @ Peraosis: — 294 km Data transmission phase (320 pixel)
[3] North pole: 2000 km
While the mission would already have been considered
highly successful had it merely managed, from August Figure 1: Schematic sketch of image acquisition by
1990 onwards, to produce single images of 70% of the Magellan's SAR sensor, including distances to the
surface, in fact it produced data also during two additional surface of the planet at distinct points
acquisition cycles: Cycle 2 was collecting images at an
opposite viewing direction and Cycle 3 provided stereo- The US$ 1 billion spent on Magellan primarily served the
Coverage in combination with Cycle 1 data by using a goal of collecting synthetic aperture radar (SAR) images
steeper inclination angle of the radar beam. Thus, many of the planet’s surface, to assemble all images to a
areas on Venus were imaged at least three times, and a mosaicked presentation on film at reduced scale, and to
Coverage much higher than that specified as a mission goal develop other data sets such as from altimetry, radar
could be achieved. backscatter and gravity operations.
491
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996
