Title
Intrinsic Dimensionality in Combined Visible to Thermal Infrared Imagery
Abstract
Intrinsic dimensionality (ID) provides an objective metric with which to quantify the number of detectable signal components in a spectroscopic image. Here, we use ID to illustrate the information gained by fusing spectroscopic data acquired over different wavelength ranges. For Cuprite, a mineral-rich site in the Nevada desert, the signal content from visible to short-wave infrared (VSWIR) describes almost entirely different signal content from the thermal infrared (TIR). Due to the extremely limited number of coincident VSWIR and TIR acquisitions previously acquired, this article provides a unique opportunity to quantify the information content gained by adding TIR acquisitions to the more commonly acquired VSWIR data. We highlight the importance of combined VSWIR/TIR imaging for the complete characterization and mapping of mineral and other sites.
Year
DOI
Venue
2019
10.1109/JSTARS.2019.2938883
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Keywords
Field
DocType
Minerals,Remote sensing,Covariance matrices,Eigenvalues and eigenfunctions,Rocks,Image resolution,Earth
Computer vision,Thermal infrared,Remote sensing,Curse of dimensionality,Artificial intelligence,Mathematics,Instrumental and intrinsic value
Journal
Volume
Issue
ISSN
12
12
1939-1404
Citations 
PageRank 
References 
0
0.34
0
Authors
4
Name
Order
Citations
PageRank
Kerry Cawse-Nicholson102.70
Simon J. Hook214688.70
Charles E. Miller3175.56
David Ray Thompson400.34