Statistics – Applications
Scientific paper
Jan 1999
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=1999nvm..conf...52r&link_type=abstract
Workshop on New Views of the Moon 2: Understanding the Moon Through the Integration of Diverse Datasets, p. 52
Statistics
Applications
Elevation, Imagery, Lunar Topography, Selenology, Selenography, Moon, Lunar Geology, Lunar Maps, Lunar Surface, Applications Programs (Computers), Cameras, Clementine Spacecraft, Computer Systems Programs, Laser Altimeters
Scientific paper
During the Clementine Mission both oblique and vertical multispectral images were collected. The oblique and vertical images from a single spectral band collected during the same orbit form a stereo pair that can be used to derive the topography. These stereo pairs are being used to derive the topography of an area (90 deg S to 650S latitude) surrounding the lunar south pole. Work on the lunar north pole topography will start after completion of the south pole topography. This report provides an update on the initial results for the lunar south pole topography. In 1994, the Clementine spacecraft acquired digital images of the Moon at visible and near-infrared wavelengths. Onboard there were four camera systems and a laser altimeter. During the first pass, periapsis was at 30S and the highest resolution images were obtained in the southern hemisphere. Over the northern polar area, a series of oblique and vertical images were obtained with the ultraviolet-visible (UV-VIS) camera on each orbit. During the second pass, periapsis was at 30N and the image acquisition strategy was reversed. The UV-VIS camera image size was 384 x 288 pixels with five spectral bands and one broad band. The 750-nm-band stereo pairs are the primary image source for this study. The ground sample distances (GSD) for oblique images range from 300 to 400m. The GSD for the vertical images, acquired at the end of an orbit, are slightly larger and range from 325 to 450 m. Using the formula for stereo-height accuracy, an estimate of height accuracy is 180m. This formula is IFOVMAX)/(K*B/H with IFOVMAX defined as Maximum Instantaneous Field of View; B/H is the base-to-height ratio and K is an estimate of pixel measurement accuracy on the imagery. The Clementine laser altimeter (LIDAR) data were used previously to produce a global topographic model of the Moon . The model has a vertical accuracy of about 100 m and a spatial resolution of 2.5 deg. Altimetry data were collected between 79S and 810N. These data were filtered and then interpolated to fill in the polar regions where the altimeter did not collect data. A global topography model was then derived based on spherical harmonic expansion. Image mosaic. A global image mosaic of the Moon was produced from the 750-nm. Clementine data . The mosaic includes high-resolution, oblique, and vertical images. Match points were picked to tie the imagery together, and the camera pointing angles were adjusted to align the imagery. This adjustment used a spherical surface, and the elevation of all points was held to a constant value, 1737.4 km. This produced a seamless image mosaic with latitude and longitude information but no information on the elevation. The imagery and support information were downloaded to our digital photogrammetric workstation from the Integrated Software for Imagers and Spectrometers (ISIS) system. The support data included the camera location and pointing angles. Match points used to, produce the image mosaic were also downloaded. The camera angles were adjusted to account for the elevation of the match points. This was accomplished with the Multi Sensor Triangulation (MST) software from LH Systems SOCET Set software package. The revised camera angles allowed for the derivation of a digital elevation model (DEM) from the stereo pairs. The match-point latitude and longitude from the global image mosaic are accurate and used for an initial estimate of the horizontal position. The elevations of the match points were estimated from the altimetry data. The camera angles used in the altimetry processing and in the creation of the image mosaic were adjusted independently. Hence, the horizontal position of the altimetry data and the image mosaic are not aligned correctly. Clementine was designed so the altimeter shared the optical system of the HIRES camera system. The HIRES and UV-VIS camera systems were aligned so the HIRES image was centered in the UV-VIS image. We therefore made an adjustment so that the altimetry points would fall near the centerline of the UV-VIS imagery. A DEM was created from the altimetry data using the adjusted position and the match-point elevations were estimated from the adjusted DEM. These points were used in the HATS software, which allows for the use of weights on the estimated position based on the accuracy of the point. The horizontal positions were given a weight of 1 km, and the vertical estimates were given a weight of 5 km. In forming the Clementine Mosaic, over 3600 images and 29,000 match points were used in the southern polar region, an area defined as 64S or less. Different techniques were tried in adjusting the images and match points. Limitations of computer memory required breaking the data into blocks that would be solved separately and then brought back together. Initial results had large elevation errors in the DEMs derived from different stereo pairs. This was most likely caused by the fact that although the stereo models overlapped, they did not share similar match points. One solution for this problem would require transferring all 29,000 match points to all the images they fall on within the set of 36M images. This would be a tremendous amount of work. The solution used was to thin the images to just the stereo pairs (983 images) and to thin the match points to cover the area (973 points.) The MST software was then used to add match points with the criteria that each image should have nine match points distributed throughout the image. This process added 1181 points. During editing and checking the match points, 45 more match points were added. The number of match points within an image ranges from 3 to 30 points, with 96% of the images having nine or more points and the average image having 16 points. Additional information contained in original.
Howington-Kraus A.
Kirk Randolph
Rosiek Mark R.
No associations
LandOfFree
Lunar South Pole Topography Derived from Clementine Imagery does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Lunar South Pole Topography Derived from Clementine Imagery, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Lunar South Pole Topography Derived from Clementine Imagery will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1276563