HR: 0830h
AN: P41B-0410    [PDF]
TI: Automated geologic mapping in Hellas Basin, Mars using THEMIS and GENIE
AU: * Kraal, E R
EM: ekraal@es.ucsc.edu
AF: University of California Santa Cruz, Department of Earth Science 1156 High Street, Santa Cruz, CA 95064 United States
AU: Brumby, S P
AF: Los Alamos National Laboratory, Space and Remote Sensing Sciences Mail Stop D436, Los Alamos, NM 87545 United States
AU: Asphaug, E I
AF: University of California Santa Cruz, Department of Earth Science 1156 High Street, Santa Cruz, CA 95064 United States
AB: Most of Mars has been geological mapped using traditional methods of visual interpretation of images from orbiting spacecraft. New high spatial resolution multispectral data being returned by THEMIS contains valuable, additional information that can be compared to and used to improve these traditional maps. Automating the analysis of large amounts of multispectral imagery requires new, adaptive software tools. We use the Los Alamos GENIE (GENetic Imagery Exploitation) genetic programming system to generate automated spatio-spectral image processing algorithms for comparing and mapping the martian surface using THEMIS multispectral images. Our test area is the Hellas Basin because it has good coverage in the released THEMIS images. In addition, it has been extensively mapped and presents a variety of geologic features: impact craters, aeolian deposits, volcanics, lacustrine/ice features, and channels representing a broad spectrum of geologic units on Mars. We present results that compare previous geologic mapping of the Hellas Basin to new compositional spectral maps. In addition, we test automatic geologic mapping guided by limited human input, and discuss how this approach could enable a more efficient process for analyzing multispectral imagery.
DE: 5464 Remote sensing
DE: 5470 Surface materials and properties
DE: 6225 Mars
SC: Planetary Sciences [P]
MN: 2003 Fall Meeting