Performance of algorithms for target signal detection in Hyperspectral Imagery (HSI) is often deteriorated when the data is neither statistically homogeneous nor Gaussian or when its Joint Probability Density (JPD) does not match any presumed particular parametric model. In this paper we propose a novel detection algorithm which first attempts at dividing data domain into mostly Gaussian and mostly Non-Gaussian (NG) subspaces, and then estimates the JPD of the NG subspace with a non-parametric Graph-based estimator. It then combines commonly used detection algorithms operating on the mostly-Gaussian sub-space and an LRT calculated directly with the estimated JPD of the NG sub-space, to detect anomalies and known additive-type target signals. The algorithm performance is compared to commonly used algorithms and is found to be superior in some important cases.

Date Published: 29 May 2013
PDF: 11 pages
Proc. SPIE 8743, Algorithms and Technologies for Multispectral, Hyperspectral, and Ultraspectral Imagery XIX, 87431A (29 May 2013); doi: 10.1117/12.2016452

Published in SPIE Proceedings Vol. 8743:
Algorithms and Technologies for Multispectral, Hyperspectral, and Ultraspectral Imagery XIX
Sylvia S. Shen; Paul E. Lewis, Editor(s)