Engineering topics

Gilmore, Martha S.

Publications and source records attributed to Gilmore, Martha S..

7 records

NASA NTRS2013

Metric Learning to Enhance Hyperspectral Image Segmentation

Unsupervised hyperspectral image segmentation can reveal spatial trends that show the physical structure of the scene to an analyst. They highlight borders and reveal areas of homogeneity and change. Segmentations are independently helpful for object recognition, and assist with automated production of symbolic maps. Additionally, a good segmentation can dramatically reduce the number of effective spectra in an image, enabling analyses that would otherwise be computationally prohibitive. Specifically, using an over-segmentation of the image instead of individual pixels can reduce noise and potentially improve the results of statistical post-analysis. In this innovation, a metric learning approach is presented to improve the performance of unsupervised hyperspectral image segmentation. The prototype demonstrations attempt a superpixel segmentation in which the image is conservatively over-segmented; that is, the single surface features may be split into multiple segments, but each individual segment, or superpixel, is ensured to have homogenous mineralogy.

Thompson, David R.↗

NASA NTRS2011

Metric Learning for Hyperspectral Image Segmentation

We present a metric learning approach to improve the performance of unsupervised hyperspectral image segmentation. Unsupervised spatial segmentation can assist both user visualization and automatic recognition of surface features. Analysts can use spatially-continuous segments to decrease noise levels and/or localize feature boundaries. However, existing segmentation methods use tasks-agnostic measures of similarity. Here we learn task-specific similarity measures from training data, improving segment fidelity to classes of interest. Multiclass Linear Discriminate Analysis produces a linear transform that optimally separates a labeled set of training classes. The defines a distance metric that generalized to a new scenes, enabling graph-based segmentation that emphasizes key spectral features. We describe tests based on data from the Compact Reconnaissance Imaging Spectrometer (CRISM) in which learned metrics improve segment homogeneity with respect to mineralogical classes.

Compact Reconnaissance Imaging Spectrometer (CRISM↗

NASA NTRS2009

Sparse Superpixel Unmixing for Exploratory Analysis of CRISM Hyperspectral Images

Fast automated analysis of hyperspectral imagery can inform observation planning and tactical decisions during planetary exploration. Products such as mineralogical maps can focus analysts' attention on areas of interest and assist data mining in large hyperspectral catalogs. In this work, sparse spectral unmixing drafts mineral abundance maps with Compact Reconnaissance Imaging Spectrometer (CRISM) images from the Mars Reconnaissance Orbiter. We demonstrate a novel "superpixel" segmentation strategy enabling efficient unmixing in an interactive session. Tests correlate automatic unmixing results based on redundant spectral libraries against hand-tuned summary products currently in use by CRISM researchers.

Image Segmentation↗

NASA NTRS2005

Creation and testing of an artificial neural network based carbonate detector for Mars rovers

We have developed an artificial neural network (ANN) based carbonate detector capable of running on current and future rover hardware. The detector can identify calcite in visible/NIR (350-2500 nm) spectra of both laboratory specimens covered by ferric dust and rocks in Mars analogue field environments. The ANN was trained using the Backpropagation algorithm with sigmoid activation neurons. For the training dataset, we chose nine carbonate and eight non-carbonate representative mineral spectra from the USGS spectral library. Using these spectra as seeds, we generated 10,000 variants with up to 2% Gaussian noise in each reflectance measurement. We cross-validated several ANN architectures, training on 9,900 spectra and testing on the remaining 100. The best performing ANN correctly detected, with perfect accuracy, the presence (or absence) of carbonate in spectral data taken on field samples from the Mojave desert and clean, pure marbles from CT. Sensitivity experiments with JSC Mars-1 simulant dust suggest the carbonate detector would perform well in aeolian Martian environments.

neural networks↗

NASA NTRS2004

Effect of Palagonite Dust Deposition on the Automated Detection of Carbonate Vis/NIR Spectra

Currently Mars missions can collect more data than can be returned. Future rovers of increased mission lifetime will benefit from onboard autonomous data processing systems to guide the selection, measurement and return of scientifically important data. One approach is to train a neural net to recognize spectral reflectance characteristics of minerals of interest. We have developed a carbonate detector using a neural net algorithm trained on 10,000 synthetic Vis/NIR (350-2500 nm) spectra. The detector was able to correctly identify carbonates in the spectra of 30 carbonate and noncarbonate field samples with 100% success. However, Martian dust coatings strongly affect the spectral characteristics of surface rocks potentially masking the underlying substrate rock. In this experiment, we measure Vis/NIR spectra of calcite coated with different thicknesses of palagonite dust and evaluate the performance of the carbonate detector.

Gilmore, Martha S.↗

NASA NTRS2004

Depths and Geologic Setting of Northern Hemisphere Gullies (and Comparison to Their Southern Counterparts)

Gullies are recognized as having an incised (often sinuous) channel, alcove, and apron morphology as defined. Previous detailed work has focused on the characteristics of southern hemisphere gullies; here we examine the distribution, morphology and depth of northern hemisphere gullies. Gully locations (both hemispheres) were compiled from the literature, not a global search, and thus the gullies studied here are the subset of the whole for which corresponding MOLA tracks are available. We find that the heads of the majority of N. hemisphere gullies commence in a specific cliff-forming layer, and average 250 m below the local surface. This observation suggests that the rock layer has a genetic relationship to gullies and is difficult at present to reconcile with gully formation from snowmelt, which predicts no correlation to rock layers. Gully locations and depths correspond to specific geologic units. We propose that in addition to ground ice and a melting mechanism, gully formation requires groundwater flow along or within specific rock units.

Gilmore, Martha S.↗

NASA NTRS1999

Venus Surface Sample Return: A Weighty High-Pressure Challenge

A mission to return a sample to Earth from the surface of Venus faces a multitude of multidisciplinary challenges. In addition to the complications inherent in any sample return mission, Venus presents the additional difficulties of a deep gravity well essentially equivalent to Earth's and a hot-house atmosphere which generates extremes of high temperature, density, and pressure unmatched at any other known surface in the solar system. The Jet Propulsion Laboratory of the California Institute of Technology recently conducted a study to develop an architecture for such a mission; a major goal of this study was to identify technology developments which would need to be pursued in order to make such a mission feasible at a cost much less than estimated in previous. The final design of this mission is years away but the study results presented here show our current mission architecture as it applies to a particular mission opportunity, give a summary of the engineering and science trades which were made in the process of developing it, and identify the main technology development efforts needed.

Sweetser, Ted↗