|
|
|
Autonomous Mineral Detectors for Mars Rovers and Landers
|
|
|
| We will design software that will provide instrumented rovers the ability to select and analyze spectroscopic data, characterize a landing site and test geological hypotheses autonomously. Our approach is to combine supervised methods for focused mineral detection and broader mineral classification with an unsupervised clusterer that identified unusual or previously unencountered mineral classes. We will 1) improve our current carbonate detector by validating its sensitivity and performing more detailed modeling of nonlinear mixture behavior observed in intimate mineral mixtures, 2) identify other mineral classes of interest, particularly those known to exist on Mars and be indicators of life on Earth, and design detectors for those minerals, and 3) develop a novelty detector that can identify new minerals in a series of observations. The algorithms will be trained using synthetic data from spectral libraries and tested in real data collected in the field. |
Bibliography
| | Creation and tecting of an artificial neural network based carbonate detector for Mars rovers
Bornstein, Benjamin; Castano, R.; Gilmore, M. S.; Merrill M. and Greenwood J. P., IEEE Aerospace Conf., IEEEAC Paper #1527
| |
| EFFECT OF PALAGONITE DUST DEPOSITION ON THE AUTOMATED DETECTION OF CARBONATE VIS/NIR SPECTRA
GILMORE M. S.; *MERRILL M. D.; CASTAÑO R.; BORNSTEIN B. AND GREENWOOD J., Lunar and Planetary Science Conference, (2004) 35 pp. Abstract #1335 (CD-ROM)
| |
| Effect of Mars analogue dust deposition on the automated detection of calcite in visible/near-infrared spectra
Gilmore, Martha S.; Merrill, M. D.; Castano, R.; BORNSTEIN B. AND GREENWOOD J., Icarus, 172 pp. 641-646
| |
|
|
