Adapting to Change: The CFAR Problem in Advanced Hyperspectral Detection

Newer, realistic models of targets and backgrounds used in hyperspectral detection do not always lend themselves to a CFAR (constant false alarm rate) formulation. Several advanced techniques are considered here. It is found that incorporating a particular empirically validated method of target evolution permits an exact CFAR version of a large class of advanced detectors based on elliptically contoured distributions. Other validated detectors are considered, for which no closed form normalization exists to convert them to CFAR form. For these a geometrical approach to achieving approximate CFAR performance is described and analyzed.

[1]  A. Schaum,et al.  Linear chromodynamics models for hyperspectral target detection , 2003, 2003 IEEE Aerospace Conference Proceedings (Cat. No.03TH8652).

[2]  Eric Allman,et al.  Hyperspectral change detection in high clutter using elliptically contoured distributions , 2007, SPIE Defense + Commercial Sensing.

[3]  Alan P. Schaum.  Hyperspectral detection algorithms: operational, next generation, on the horizon , 2005, 34th Applied Imagery and Pattern Recognition Workshop (AIPR'05).

[4]  A. Schaum.  ADVANCED HYPERSPECTRAL ALGORITHMS FOR TACTICAL TARGET DETECTION AND DISCRIMINATION , 2008 .