Share Email Print
cover

Proceedings Paper

DMD based hyperspectral augmentation for multi-object tracking systems
Author(s): Jonathan G. Neumann
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Current wide area pan-chromatic or dual band persistent surveillance systems often do not provide enough observable data to maintain accurate and unique tracks of multiple objects over a large field of regard. Previous experiments have shown augmentation with hyperspectral imagery can enhance tracking performance. However, hyperspectral imagers have significantly slower coverage rates than persistent surveillance systems, essentially because a hyperspectral pixel contains vector data while a standard image pixel is typically either scalar or RGB. This coverage rate gap is a fundamental mismatch between the two systems and presents a technological hurdle to the practical use of hyperspectral imagers for tracking multiple objects spread over the entire field of regard of persistent surveillance systems. In this non-mapping hyperspectral application, we assume much of the information in a hyperspectral data cube is superfluous background and need not be processed or even collected. The effective coverage rate of the hyperspectral imager can be made compatible with modern persistent surveillance systems, at least for the objects of interest, by using a DMD to judiciously select which data are collected and processed. A proof-of-concept sensor has been developed and preliminary results are presented.

Paper Details

Date Published: 14 February 2009
PDF: 12 pages
Proc. SPIE 7210, Emerging Digital Micromirror Device Based Systems and Applications, 72100B (14 February 2009); doi: 10.1117/12.811551
Show Author Affiliations
Jonathan G. Neumann, Naval Research Lab. (United States)


Published in SPIE Proceedings Vol. 7210:
Emerging Digital Micromirror Device Based Systems and Applications
Larry J. Hornbeck; Michael R. Douglass, Editor(s)

© SPIE. Terms of Use
Back to Top