Spectral Intelligence™

OSK PLATFORMS

HYPERSPECTRAL DATA CUBE

OSK SPECTRAL INTELLIGENCE™

  • Each pixel provides hundreds of data points for chemical ID
  • Chemical signatures can be matched to ID elements, compounds & objects
  • Algorithms optimized for rapid object detection/classification

What is hyperspectral? What is Spectral Intelligence™? Current optical imagery space-platforms only capture a handful of wide color bands in the visible spectrum and rely on high spatial resolution or temperature data, which severely limits their ability to distinguish objects, materials, and chemical compounds. Whereas our platform captures hundreds of narrow contiguous color bands in the visible and infrared spectra, providing a unique chemical fingerprint of each target in our data set. With Orbital Sidekick’s Spectral Intelligence™, we can effectively see the unseen.

WHAT SETS OSK APART – SPECTRAL INTELLIGENCE™

Current remote sensing technology can tell you if you’re looking at a pipeline. We can tell you if a pipeline is leaking, and not only that it’s leaking but WHAT it is leaking.

With our Spectral Intelligence™ platform, we can dive into the future of change detection from a spectral standpoint as opposed to the outdated ways of spatial object detection. We can apply computationally efficient change detection and supervised classification machine learning algorithms to detect differences in the spectral signatures of each pixel as a function of time, and from there we can correlate these changes to our ever-expanding spectral library to chemically classify the basis for that change.

OSK’s unmatched and unparalleled spectral resolution enables monitoring of chemical change with 20X greater sensitivity than current monitoring systems at 1/10th of the cost. This daily monitoring of chemical change on a local and global scale will replace ineffective, invasive, and costly ground-based and aerial sensor systems.

How Spectral Intelligence™ Works

OSK platforms capture hyperspectral data and transmit to OSK headquarters

1

New datasets for specific targets
are compared to historical
datasets

2

Hundreds of color bands and
reflectance features are analyzed to
determine change and anomalies
at a chemical level

3

Chemical makeup of each pixel is
indexed according to a high fidelity
reference library for classification
and speciation

4

Monitoring information delivered to
customer’s existing database
architecture