Anomaly detection from hyperspectral imagery

By D. Stein, S. Beaven, L. Hoff, E. M. Winter, Alan P. Schaum, A. Stocker. IEEE Signal Processing Magazine, 2002

756 citations recorded by the source corpus.

We develop anomaly detectors, i.e., detectors that do not presuppose a signature model of one or more dimensions, for three clutter models: the local normal model, the global normal mixture model, and the global linear mixture model. The local normal model treats the neighborhood of a pixel as having a normal probability distribution. The normal mixture model considers the observation from each pixel as arising from one of several possible classes such that each class has a normal probability distribution. The linear mixture model considers each observation to be a linear combination of fixed spectra, known as endmembers, that are, or may be, associated with materials in the scene, and the coefficients, interpreted as fractional abundance, are constrained to be nonnegative and sum to one. We show how the generalized likelihood ratio test (GLRT) may be used to derive anomaly detectors for the local normal and global normal mixture models. The anomaly detector applied with the linear mixture approach proceeds by identifying target like endmembers based on properties of the histogram of the abundance estimates and employing a matched filter in the space of abundance estimates. To overcome the limitations of the individual models, we develop a joint decision logic, based on a maximum entropy probability model and the GLRT, that utilizes multiple decision statistics, and we apply this approach using the detection statistics derived from the three clutter models. Examples demonstrate that the joint decision logic can improve detection performance in comparison with the individual anomaly detectors. We also describe the application of linear prediction filters to repeated images of the same area to detect changes that occur within the scene over time.

Subfields: Artificial Intelligence, Atmospheric Science, Media Technology

View Paper: https://www.semanticscholar.org/paper/95a97cd4d7ecd8b8a2d84577eebb5541077f46aa

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Atmospheric Science

In Media Technology

Citations in the corpus

Hyperspectral Unmixing Overview: Geometrical, Statistical, and Sparse Regression-Based Approaches (2012, neutral): “H YPERSPECTRAL cameras [1]–[11] contribute significantly to earth observation and remote sensing [12], [13].”
Hyperspectral Image Classification—Traditional to Deep Models: A Survey for Future Prospects (2021, background): “Thus, considering the aforementioned limitations, HSI analysis is categorized into the following main streams: dimensionality reduction [6]–[8], spectral unmixing [9]–[14], change detection [15]–[17] classification [18]–[21], feature learning for classification [22]–[24], restoration and denoising [”
Hyperspectral Image Processing for Automatic Target Detection Applications (2003, background): “Such stochastic mixing models have been developed elsewhere [ 7 ], but their applicability is limited because the estimation of their parameters leads to highly nonlinear optimization problems.”
Transform Coding Techniques for Lossy Hyperspectral Data Compression (2007, background): “Detectors that do not require target signature models have been developed [36], so as to overcome the difficulty of developing such signatures; such algorithms are usually referred to as " anomaly detectors.”
Nonnegative Least-Correlated Component Analysis for Separation of Dependent Sources by Volume Maximization (2010, background): “, s 1⁄2n 1N 1⁄4 1 for alln) in hyperspectral imaging [29].”
Hyperspectral Remote Sensing Data Analysis and Future Challenges (2013, methodology): “The first stage is an anomaly detector [123], [124] which identifies spectral”
A Low-Rank and Sparse Matrix Decomposition-Based Mahalanobis Distance Method for Hyperspectral Anomaly Detection (2016, background): “the results are often unstable when the local background contains large disturbances or multiple classes [10], [11].”
Modern Trends in Hyperspectral Image Analysis: A Review (2018, background): “Anomaly detection methods [110]–[112] that use spectral information to differentiate between targets and background without any prior knowledge have gained significant popularity over the past few years.”
Kernel RX-algorithm: a nonlinear anomaly detector for hyperspectral imagery (2005, methodology): “Spectral anomaly detection algorithms [8], [10]–[13] could also use spectral signatures to detect anomalies embedded within background clutter with a very low signal-to-noise ratio (SNR).”
A tutorial overview of anomaly detection in hyperspectral images (2010, methodology): “In this way, the mixture pdf fxIHo(!.) in (10 ) is completely specified, and then the GLRT can be applied [ 72 , 13].”
A support vector method for anomaly detection in hyperspectral imagery (2006, background): “In many environments, it has been shown empirically that the local normal model provides an inadequate representation of the underlying distribution [9] leading to poor false alarm performance.”