Share Email Print

Proceedings Paper

Subpixel detection of surface mines in hyperspectral images
Author(s): Glenn E. Healey; David Slater
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Hyperspectral signatures for surface mines in airborne images can have substantial variability due to the environmental conditions and subpixel mixing. Signatures are also affected by the condition of the mine. We show that a subspace representation for mine spectral properties can be used as the basis for an algorithm for subpixel mine detection that is invariant to the illumination and atmospheric conditions. A background model is estimated from the image data to support subpixel detection. The intrinsic spectral reflectance of the mine is the only input required by the algorithm. We demonstrate the performance of the algorithm for several mine types over a range of conditions and altitudes in visible through near-infrared hyperspectral images. Several of the mine types appear at a scale that is significantly smaller than a pixel.

Paper Details

Date Published: 21 September 2004
PDF: 11 pages
Proc. SPIE 5415, Detection and Remediation Technologies for Mines and Minelike Targets IX, (21 September 2004); doi: 10.1117/12.542344
Show Author Affiliations
Glenn E. Healey, HyperTech Systems, LLC (United States)
David Slater, HyperTech Systems, LLC (United States)

Published in SPIE Proceedings Vol. 5415:
Detection and Remediation Technologies for Mines and Minelike Targets IX
Russell S. Harmon; J. Thomas Broach; John H. Holloway Jr., Editor(s)

© SPIE. Terms of Use
Back to Top