Intrinsic Dimensionality in Combined Visible to Thermal Infrared Imagery

Cawse-Nicholson, K., S. Hook, C. E. Miller, and D. Thompson (2019), Intrinsic Dimensionality in Combined Visible to Thermal Infrared Imagery, IEEE Journal Of Selected Topics In Applied Earth Observations And Remote Sensing, 12, 4977-4984, doi:10.1109/JSTARS.2019.2938883.

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.

PDF of Publication: 
Download from publisher's website.