Hyperspectral Image (HSI) anomaly detectors typically employ local background modeling techniques to facilitate target
detection from surrounding clutter. Global background modeling has been challenging due to the multi-modal content
that must be automatically modeled to enable target/background separation. We have previously developed a support
vector based anomaly detector that does not impose an a priori parametric model on the data and enables multi-modal
modeling of large background regions with inhomogeneous content. Effective application of this support vector
approach requires the setting of a kernel parameter that controls the tightness of the model fit to the background data.
Estimation of the kernel parameter has typically considered Type I / false-positive error optimization due to the
availability of background samples, but this approach has not proven effective for general application since these
methods only control the false alarm level, without any optimization for maximizing detection. Parameter optimization
with respect to Type II / false-negative error has remained elusive due to the lack of sufficient target training exemplars.
We present an approach that optimizes parameter selection based on both Type I and Type II error criteria by
introducing outliers based on existing hypercube content to guide parameter estimation. The approach has been applied
to hyperspectral imagery and has demonstrated automatic estimation of parameters consistent with those that were found
to be optimal, thereby providing an automated method for general anomaly detection applications.
Existing techniques for hyperspectral image (HSI) anomaly detection are computationally intensive precluding real-time
implementation. The high dimensionality of the spatial/spectral hypercube with associated correlations between spectral
bands present significant impediments to real time full hypercube processing that accurately encapsulates the underlying
modeling. Traditional techniques have imposed Gaussian models, but these have suffered from significant
computational requirements to compute large inverse covariance matrices as well as modeling inaccuracies. We have
developed a novel data-driven, non-parametric HSI anomaly detector that has significantly reduced computational
complexity with enhanced HSI modeling, providing the capability for real time performance with detection rates that
match or surpass existing approaches. This detector, based on the Support Vector Data Description (SVDD), provides
accurate, automated modeling of multi-modal data, facilitating effective application of a global background estimation
technique which provides the capability for real time operation on a standard PC platform. We have demonstrated one
second processing time on hypercubes of dimension 256×256×145, along with superior detection performance relative to
alternate detectors. Computation performance analysis has been quantified via processing runtimes on a PC platform,
and detection/false-alarm performance is described via Region Operating Characteristic (ROC) curve analysis for the
SVDD anomaly detector vs. alternate anomaly detectors.
Hyper-spectral imagery (HSI) contains significant spectral resolution that enables material identification. Typical methods of classification include various forms of matching sample image spectra to pure end-member sample spectra or mixtures of these end-members. Often, pure end-members are not available a-priori. We propose the use of HSI to complement other sensor modalities which are used to cue the end-member selection process for target detection. Multiple sensor modalities are frequently available and sensor fusion is exploited as demonstrated by the DARPA Dynamic Database (DDB) and Multisensor Exploitation Testbed (MSET) programs. Candidate target pixels, cued from other sensor modalities, are registered to the HSI and verified using local matched filters. Target identification is then performed using multiple methods including Euclidean distance, spectral angle mapping, anomaly detection, principal component analysis (PCA) decomposition and reconstruction, and linear discriminant analysis (LDA). The use of LDA for target identification as well as scene segmentation provides significant capabilities to HSI understanding.
Recognition of targets in synthetic aperture radar (SAR) imagery is approached from the viewpoint of an optimization problem. Features are extracted from SAR target images and are treated as point sets. The matching problem is formulated as a non-linear objective function to maximize the number of matched features and minimize the distance between features. The minimum of this function is found using a deterministic annealing process. Registration is performed iteratively by using an analytically computed minimum at each temperature of the annealing. Thus, the images do not need to be initially registered as any translational error between them is solved for as part of the optimization. We have also extended the initial objective function to incorporate multiple feature classes. This matching method is robust to spurious, missing and migrating features. Matching results are presented for simulated XPATCH and real MSTAR SAR target imagery demonstrating the utility of this approach.
This paper discusses algorithms that are useful for the classification of targets in SAR imagery. Two algorithms are presented for segmenting a target region from background clutter; one based on constant false alarm rate detection, and another histogram based technique. The histogram based technique is extended to extract shadow regions associated with a target. A method is then presented for estimating the orientation of segmented targets. These algorithms are applied to SAR imagery from the Lincoln Lab ADTS and MSTAR datasets. The aspect estimate is shown to be superior to estimates obtained from the direction of the axis of least inertia.
Automatic classification of target in synthetic aperture radar (SAR) imagery is performed using topographic features. Targets are segmented from wide area imagery using a constant false alarm rate (CFAR) detector. Individual target areas are classified using the topographical primal sketch which assigns each pixel a label that is invariant under monotonic gray tone transformations. A local surface fit is used to estimate the underlying function oat each target pixel. Pixels are classified based on the zero crossings of the first directional derivatives and the extrema of second directional derivatives. These topographic labels along with the quantitative values of second directional derivative extrema and gradient are used in target matching schemes. Multiple matching schemes are investigated including correlation and graph matching schemes that incorporate distance between features as well as similarity measures. Cost functions are tailored to the topographic features inherent in SAR imagery. Trade offs between the different matching schemes are addressed with respect to robustness and computational complexity. Classification is performed using one foot and one meter imagery obtained from XPATCH simulations and the MSTAR synthetic dataset.
We address the application of model-supported exploitation techniques to synthetic aperture radar (SAR) imagery. The emphasis is on monitoring SAR imagery using wide area 2D and/or 3D site models along with contextual information. We consider here the following tasks useful in monitoring: (a) site model construction using segmentation and labeling techniques, (b) target detection, (c) target classification and indexing, and (d) SAR image-site model registration. The 2-D wide area site models used here for SAR image exploitation differ from typical site models developed for RADIUS applications, in that they do not model specific facilities, but constitute wide area site models of cultural features such as urban clutter areas, roads, clearings, fields, etc. These models may be derived directly from existing site models, possibly constructed from electro-optical (EO) observations. When such models are not available, a set of segmentation and labeling techniques described here can be used for the construction of 2D site models. The use of models can potentially yield critical information which can disambiguate target signatures in SAR images. We address registration of SAR and EO images to a common site model. Specific derivations are given for the case of registration within the RCDE platform. We suggest a constant false alarm rate (CFAR) detection scheme and a topographic primal sketch (TPS) based classification scheme for monitoring target occurrences in SAR images. The TPS of an observed target is matched against candidate targets TPSs synthesized for the preferred target orientation, inferred from context (e.g. road or parking lot targets). Experimental results on real and synthetic SAR images are provided.
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.