Anomaly Detection in Hyperspectral Imagery

Abstract

In this chapter we are presenting the literature and proposed approaches for anomaly detection in hyperspectral images. These approaches are grouped into four categories based on the underlying techniques used to achieve the detection: 1) the statistical based methods, 2) the kernel based methods, 3) the feature selection based methods and 4) the segmentation based methods. Since the first approaches are mostly based on statistics, the recent works tend to be more geometrical or topological especially with high resolution images where the high resolution implies the presence of many materials in the same geographic area that cannot be easily distinguished by usual statistical methods.

Random Subspace Mixture Models for Interpretable Anomaly Detection

“…Mixture models have a long history of being used to detect anomalies [30,12,31,15,14] for both numerical and categorical data [2]. Various techniques have been used to determine the number of components for a mixture model such as greedy EM-learning [32], minimal message length [20], stacking [22], holdout techniques [33], and assumptions on the components distribution [22].…”

“…It includes both outlier detection and novelty detection and is a field of study that has been around for a long time [13]. Anomaly detection has been used to find irregularities in many different domains including in hyperspectral imagery [14], maritime surveillance [15], healthcare analytics [16], medical applications [17], and the Internet of Things [18]. It is also used for fraud detection [19], intrusion detection [20], flight operation and safety monitoring [21], and to detect misinformation [22].…”