The Projection-Slice Synthetic Discriminant Function Filter has been generated using a sparse sampling technique that utilizes the inherent sprsity of the Projection-Slice theorem. The l1-norm has been utilized to optimize the information contents extracted from the representative class objects. In this work, the results of the usual PSDF without the benefit of convex optimization is compared with the results of the PSDF filter after utilization of convex optimization to assess the merits of the utilization of efficient information reconstruction within the construct of the PSDF.

Date Published: 13 October 2014
PDF: 12 pages
Proc. SPIE 9250, Electro-Optical Remote Sensing, Photonic Technologies, and Applications VIII; and Military Applications in Hyperspectral Imaging and High Spatial Resolution Sensing II, 925011 (13 October 2014); doi: 10.1117/12.2059841

Published in SPIE Proceedings Vol. 9250:
Electro-Optical Remote Sensing, Photonic Technologies, and Applications VIII; and Military Applications in Hyperspectral Imaging and High Spatial Resolution Sensing II
Gary Kamerman; Gary J. Bishop; Ainsley Killey; Ove Steinvall; John D. Gonglewski, Editor(s)