Advanced ground and space-based hyperspectral imager (HSI) concepts are being developed for a wide variety of
scientific, civil, and military applications. Users and developers of these systems often require the specification of system
performance in terms of receiver-operator characteristic (ROC) curves which plot probability-of-detection (POD) versus
probability-of-false-alarm (PFA). In this paper we describe and illustrate the use of a scene-based modeling tool used to
explore ROC curve parametric dependencies on target, scene, and HSI sensor design characteristics and detection
algorithms in the visible/near infrared to shortwave infrared (VNIR/SWIR) spectral regime (i.e. from 0.4 to 2.5 microns).
The magnitudes of the target and background spectral signatures are synthesized using MODTRAN; this accounts for
pertinent solar elevation angle and albedo assumptions. Selected spectral input scenes (based on measured data) are used
assuming imbedded spectral targets (selectable), where a fill-factor parameter is used to account for target dimension
compared to sensor ground footprint. The HSI sensor sensitivity characteristics are imbedded via the noise-equivalent
reflectivity difference (NE▵ρ) figure-of-merit which is computed spectrally based on a given sensor design
configuration. Finally the POD, PFA and hence ROC parametrics are generated using a distinct candidate detection
algorithm. The roles of scene clutter, illumination conditions, and sensor signal-to-noise ratio are made clear in
simulation examples. In addition the impact of limited scene extent (limited scene pixel count) on the accuracy of the
PFA predictions is noted and discussed.
Image pixels represent either distinct materials (end members) that are present in the image, or mistures of two or more of these pure materials. Estimates of pure end member spectra are needed for spectral libraries and for input into pixel unmixing codes. We investigate three algorithms for estimating end member spectra: (1) the convex hull method in which an n-dimensional surface is shrink- wrapped around the data cloud; (2) a pixel-by-pixel search method in which pixels that have sufficiently different spectral angles are declared end members; (3) a pixel-by- pixel search method using Euclidean distance as a measure, followed by clustering to improve the estimate of the spectra. The convex hull technique should provide an estimate of pure end member spectra while the pixel-by-pixel search methods should find both distinct end members and distinct mixtures. Each method requires user-set thresholds to find distinct spectra, which can be expressed in spectral angle degrees or image-dependent units for Euclidean distance. Estimates for the lower threshold (below which two spectra are considered to be the same material) and the upper threshold (above which two spectra are definitely different materials) are derived empirically. Low-altitude AVIRIS data will be used to demonstrate the utility of these end member extraction methods. We will illusxtrate how well each technique compare to the other, and compare how well individual algorithms work across adjacent scenes.
We have developed a methodology for wavelength band selection. This methodology can be used in system design studies to provide an optimal sensor cost, data reduction, and data utility trade-off relative to a specific application. The methodology combines an information theory- based criterion for band selection with a genetic algorithm to search for a near-optimal solution. We have applied this methodology to 612 material spectra from a combined database to determine the band locations for 6, 9, 15, 30, and 60- band sets in the 0.42 to 2.5 microns spectral region that permit the best material separation. These optimal bands sets were then evaluated in terms of their utility related to anomaly ddetection and material identification using multi-band data cubes generated from two HYDICE cubes. The optimal band locations and their corresponding entropies are given in this paper. Our optimal band locations for the 6, 9, and 15-band sets are compared to the bands of existing multi-band systems such as Landsat 7, Multispectral Thermal Imager, Advanced Land Imager, Daedalus, and M7. Also presented are the anomaly detection and material identification results obtained from our generalted multi- band data cubes. Comparisons are made between these exploitation results with those obtained from the original 210-band HYDICE data cubes.
Urban areas provide a complex material environment, both in the number of materials present, and in the spatial scale of material variation. Classification in urban environments using multispectral sensors has typically been limited to discrimination of major terrain classes due to both the limited spatial resolution of currently available sensors and to the inability to consistently discriminate between similar materials. High spectral and spatial resolution imagery, such as collected with the HYDICE sensor, provides the opportunity to develop detailed material maps for urban areas, and to perform precise material discrimination for cultural objects. Referencing a comprehensive set of material spectra, this paper describes a procedure for land cover classification which can be automated and performed with little or no a-priori knowledge of objects in the scene.
This paper describes a methodology we have developed for wavelength band selection. This methodology combines an information theory-based criterion for selection with a genetic algorithm for searching for a near-optimal solution. We have applied this methodology to 302 material spectra in the Nonconventional Exploitation Factors database to determine the band locations for 7, 15, 30, and 60-band sets that permit the best material separation. These optical band sets wee also evaluated in terms of their utility related to anomaly/target detect in using multiband images generated from a hyperspectral digital imagery collection experiment image cube. The optimal band locations and their corresponding entropies are given in this paper. Also presented are the anomaly/target detection results obtained from using these optimal band sets.
This paper summarizes investigations made at The Aerospace Corporation, during the past year, in the field of hyperspectral material identification. A spectral feature metric which makes use of visible features in the `graph' of a reference spectrum or acquired hyperspectral sample was developed and refined. Using this, and other well-known spectral similarity metrics, a mechanism for creating material taxonomies using clustering techniques was developed. The taxonomies and metrics were combined to create an innovative means for identifying materials and objects in hyperspectral scenes from the HYDICE sensor.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.