Mathematics – Logic
Scientific paper
Dec 2006
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2006agufm.p23c0080b&link_type=abstract
American Geophysical Union, Fall Meeting 2006, abstract #P23C-0080
Mathematics
Logic
3252 Spatial Analysis (0500), 5480 Volcanism (6063, 8148, 8450), 6225 Mars, 8485 Remote Sensing Of Volcanoes, 9820 Techniques Applicable In Three Or More Fields
Scientific paper
Nearest Neighbor (NN) analysis is a statistical technique that is commonly used to assess spatial randomness within a feature field. We have previously applied this technique to a variety of terrestrial and Martian geological features, including volcanic rootless cones, ice mounds, impact craters, and Martian structures of unknown origin which appear as positive-relief, circular to elongate features on high-resolution images recently acquired by the Thermal Emission Imaging System (THEMIS) and the Mars Orbiter Camera (MOC). Based on the terrestrial NN results, we developed a remote sensing tool to identify classes of geologic features, and then applied the tool to make a preliminary identification of the unknown Martian features. Using this approach, we found that selected Martian feature fields have spatial distributions consistent with those of terrestrial volcanic rootless cones and/or ice mounds. Such geologic features that form in the presence of water or ice are of key interest because they provide information about the abundance and spatial distribution of volatiles. Water-rich localities on Mars not only represent potential niches for past biological development, but they are also prospective sources of water for future Mars missions. Given that the MOC and THEMIS images studied in our analysis have a range of resolutions (3-74 m/pixel), we are now taking a closer look at the NN results to assess the influence, if any, of image resolution. Our approach is to first perform NN analysis on each original (i.e., highest resolution) image. We then degrade the image by a factor of two, identify the features of interest on the degraded image, and perform the NN technique on the degraded image. We iterate this process of image degradation and NN analysis until too few features can be identified to perform the NN technique. Preliminary results obtained to date suggest that the NN results obtained at the different image resolutions are not sufficiently different as to result in a different interpretation. This suggests that the NN technique is sufficiently robust to enable comparative interpretations among feature fields which appear on images of different resolutions.
No associations
LandOfFree
A Closer Look on the Nearest Neighbor Technique to Study Geospatial Distributions: Effect of Image Resolution 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 A Closer Look on the Nearest Neighbor Technique to Study Geospatial Distributions: Effect of Image Resolution, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and A Closer Look on the Nearest Neighbor Technique to Study Geospatial Distributions: Effect of Image Resolution will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-961564