Mathematics – Logic
Scientific paper
Dec 2006
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2006agufm.p34b..02s&link_type=abstract
American Geophysical Union, Fall Meeting 2006, abstract #P34B-02
Mathematics
Logic
6225 Mars
Scientific paper
Detailed analysis of the number and morphology of impact craters on Mars provides the worth of information about the geologic history of its surface. Global catalogs of Martian craters have been compiled (for example, the Barlow catalog) but they are not comprehensive, especially for small craters. Existing methods for machine detection of craters from images suffer from low efficiency and are not practical for global surveys. We have developed a robust two-stage system for an automated cataloging of craters from digital topography data (DEM). In the first stage an innovative crater-finding transform is performed on a DEM to identify centers of potential craters, their extents, and their basic characteristics. This stage produces a preliminary catalog. In the second stage a machine learning methods are employed to eliminate false positives. Using the MOLA derived DEMs with resolution of 1/128 degrees/pixel, we have applied our system to six ~ 106 km2 sites. The system has identified 3217 craters, 43% more than are present in the Barlow catalog. The extra finds are predominantly small craters that are most difficult to account for in manual surveys. Because our automated survey is DEM-based, the resulting catalog lists craters' depths in addition to their positions, sizes, and measures of shape. This feature significantly increases the scientific utility of any catalog generated using our system. Our initial calculations yield a training set that will be used to identify craters over the entire Martian surface with estimated accuracy of 95%. Moreover, because our method is pixel-based and scale- independent, the present training set may be used to identify craters in higher resolution DEMs derived from Mars Express HRSC images. It also can be applied to future topography data from Mars and other planets. For example, it may be utilized to catalog craters on Mercury and the Moon using altimetry data to be gathered by Messenger and Lunar Reconnaissance Orbiter spacecrafts.
Mendenhall Michael P.
Stepinski Tomasz F.
No associations
LandOfFree
Robust System for Automated Identification of Martian Impact Craters 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 Robust System for Automated Identification of Martian Impact Craters, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Robust System for Automated Identification of Martian Impact Craters will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-961945