Lisa R. Gaddis
U.S. Geological Survey
A Pilot Project
Funded by the Planetary Cartography and Geologic Mapping Working Group
Planetary Geology and Geophysics Program
NASA Office of Space Science
We recently completed a Pilot Project demonstrating our capabilities for high-resolution digitization and cartographic processing with 5 full Lunar Orbiter (LO) IV frames: 109H, 110H, 114H, 115H, and 114M. Results of this project were presented at the Thirty-Second Lunar and Planetary Science Conference in March, 2001 [Gaddis et al., 2001; 1892.pdf]. These frames are approximately centered on Archimedes crater (82 km dia.) in Southeastern Imbrium Basin (Figure 1); their coverage extends from Timocharis crater in Mare Imbrium on the west to Hadley Rille (i.e., the Apollo 15 landing site) on the southeastern rim of Imbrium Basin in the east, and Aristoteles crater in Mare Frigoris the north to LaLande crater in Sinus Aestuum to the south. Geologic features in this area include Mare Imbrium, both ancient and relatively young impact craters, the Apennine Mountains, the KREEP-rich Apennine Bench Formation, Hadley Rille, Sinus Aestuum, and Rima Bode pyroclastics, Mare Vaporum, and the Imbrium basin ejecta southeast of the Imbrium Basin rim. This region provides an excellent introduction to this image product because it encompasses a wide variety of lunar geologic units, processes, and ages. These high-spatial-resolution digital data provide a very useful complement to numerous other digital geochemical, geophysical, and multispectral data for this central nearside region of the Moon.
Figure 1. Locations of 12 of the 13 Lunar Orbiter subframes
processed during the Pilot Project.
Background: The full LO dataset consists of 967 medium-resolution and 983 high-resolution frames [Bowker and Hughes, 1971]. Medium resolution views are single frames, and each high resolution frame is divided into 3 sub-frames. The LO images were photographic products acquired on the spacecraft in five NASA missions (LO I through V) while in orbit over the lunar surface in 1966 and 1967. Prior to the LO flights, the photographic film was exposed with strip numbers, a nine-level gray-scale bar, resolving power chart, and reseau marks. The original high-resolution photographic exposure was scanned into a series of strips onboard the spacecraft and then transmitted to Earth as analog data. Photographic prints from these film strips were hand mosaicked into sub-frame (for high-resolution data) and full-frame (for medium-resolution data) views and widely distributed. The resulting outstanding views were of generally very high spatial resolution (e.g., ~60 to 600 m) and covered a substantial portion of the lunar surface, but they suffered from a “venetian blind” striping, missing or duplicated data, and frequent saturation effects that hampered their use.
Data Processing: We began this work by contracting commercial digitization of selected LO positive film strips at 25-micron/pixel (e.g., ~16 m/pixel for high-resolution frames). Each scanned film strip consists of a ~16500 x 800 pixel image, with overlap between strips of ~40 pixels. We used our in-house ISIS software to process these digital data [Eliason, 1997; Gaddis et al., 1997; Torson and Becker, 1997]. Processing of the LO digital data included: (1) geometric rectification and mosaicking into subframes (3 subframes or 96 strips per frame); (2) cosmetic correction (largely destriping and noise removal); and (3) cartographic control through coregistration to the 750-nm Clementine image base [Eliason et al., 1999] via manual tie-pointing and warping (Figure 2).
Figure 2. Examples of processing stages for LO-IV subframe
110-h1. (a) Raw digital mosaic.
(b) Destriped mosaic, with excess data trimmed. (c) Geometrically rectified or warped mosaic.
(d) Mosaic with Clementine color overlay (r=750/415 nm; g=750/950; b=415/750).
Geometric rectification and mosaicking. To ensure that each film strip is aligned properly in the final mosaic, we must use the known positions of the pre-exposed reseau marks to compensate for film distortion. We identified reseau locations on each LO film strip (~35 reseaux per strip, ~2185 reseaux per high-resolution sub-frame). A weighted least-squares fit to a 1st-order polynomial describing the orthogonal positions of the reseaux was then used to determine the geometric shifts necessary to align each strip and to create a rectangular digital mosaic with no distortion (Figure 2a).
Cosmetic correction. Cosmetic correction focused largely on the removal of the prominent “venetian-blind” stripes through the application of spatial filters in ISIS [cookbook]. Film numbers and other extraneous data were first assigned to a null value, and an image of the striping patterns was created for each mosaic by applying a series of low- and high-pass spatial filters with dimensions corresponding to the stripe widths. This image of the stripe pattern is then subtracted from the mosaic, substantially improving the appearance of the mosaic (Figure 2b).
Cartographic control. The digital LO mosaics were cartographically controlled on an individual basis through coregistration to the Clementine 750-nm geodetically controlled image base via manual tie-pointing and warping with a weighted least-squares fit to a 2nd-order polynomial (Figure 2c). This process allowed us to produce LO mosaics that were colorized with an overlay of Clementine UVVIS color-ratio data (Figure 2d). We expect to use these match-point data as input to software for a LO camera model. Implementation and application of such a model will permit creation of a true cartographic product through comparison to the Clementine coordinate system for updating the LO spacecraft pointing information and evaluating optical distortion of the LO camera. To simulate such a product, we used LHZ Systems SOCET SET photogrammetric software to create a digital mosaic that is accurately tied to the Clementine 750-nm image base for Southeastern Imbrium Basin (Figure 3).
Figure 3. LO Simulated cartographic product for Southeastern
(a) LO mosaic (frames) superimposed on Clementine 750-nm image base.
(b) LO mosaic with coregistered Clementine color-ratio overlay
(r=750/415 nm; g=750/950; b=415/750).
Figure 4. Cosmetic rectification (destriping) in ISIS applied
(a) 4-110H1.jpg from the Digital Lunar Orbiter
Photographic Atlas (Gillis, 2000). (b) Results.
Acknowledgements: This pilot project was funded by the
NASA Planetary Geologic Mapping and Cartographic Working Group (M. Shepard,
Chair) of the Planetary Geology and Geophysics Program during FY99 (J.
Plescia, Task Leader). Lisa Gaddis served as Task Leader in FY00.
The following Astrogeology team members contributed significantly to this
project: Jeff Anderson, Tammy Becker, Eric Eliason, Alicyn Gitlin,
Chris Isbell, Derrick Hirsch, Annie Howington-Kraus, Randy Kirk, Ella Mae
Lee, Jeff Plescia, Janet Richie, Bob Sucharski, Tracie Sucharski, Adrienne
Wasserman, and Lynn Weller. What a team! Scientific guidance
and support from Jeff Gillis, Brad Jolliff, and Paul Spudis are very gratefully
Bowker, D.E. and J.K. Hughes, 1971, Lunar Orbiter Photographic Atlas of the Moon, NASA SP-206.
Eliason, E., 1997, Production of digital image models (DIMs) using ISIS, LPS XVIII, 331.
Eliason et al., 1999, Mission to the Moon, The Clementine UVVIS Global Mosaic, PDS CL_4001-4078.
Gaddis, L. et al., 1997, An overview of the Integrated Software for Imaging Spectrometers (ISIS), LPS XVIII, 387.
Gaddis, L. et al., 2001, Cartographic Processing of Digital Lunar Orbiter Data, in LPS XXXII, Abtract #1892, Lunar and Planetary Institute, Houston (CDROM).
Gillis, J. (ed.), 2000, Digital Lunar Orbiter Photographic Atlas of the Moon, LPI Cont. #999 (CDROM).
Torson, J. and K. Becker, 1997, ISIS---A software architecture for processing
planetary images, LPS XVIII, 1443.