Towards the Mitigation of Correlation Effects in the Analysis of Hyperspectral Imagery with Extensions to Robust Parameter Design
Abstract
Standard anomaly detectors and classifiers assume data to be uncorrelated and homogeneous, which is not inherent in Hyperspectral Imagery (HSI). To address the detection difficulty, a new method termed Iterative Linear RX (ILRX) uses a line of pixels which shows an advantage over RX, in that it mitigates some of the effects of correlation due to spatial proximity; while the iterative adaptation from Iterative Linear RX (IRX) simultaneously eliminates outliers. In this research, the application of classification algorithms using anomaly detectors to remove potential anomalies from mean vector and covariance matrix estimates and addressing non-homogeneity through cluster analysis, both of which are often ignored when detecting or classifying anomalies, are shown to improve algorithm performance. Global anomaly detectors require the user to provide various parameters to analyze an image. These user-defined settings can be thought of as control variables and certain properties of the imagery can be employed as noise variables. The presence of these separate factors suggests the use of Robust Parameter Design (RPD) to locate optimal settings for an algorithm. This research extends the standard RPD model to include three factor interactions. These new models are then applied to the Autonomous Global Anomaly Detector (AutoGAD) to demonstrate improved setting combinations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 2012
- Accession Number
- ADA565980
Entities
People
- Jason P. Williams
Organizations
- Air Force Institute of Technology