BARNARD MICROSYSTEMS LIMITED

helping you keep an eye on things...

   
     
Home / unmanned air systems / payload I / hyperspectral
 

 

   
Home | introduction | circuit design software | unmanned air systems | contact | what's new | site info
 
 

 

   

unmanned air ...

features 
issues 
applications I 
applications II 
geological survey 
oil leakage 
engines 
avionics 
airframes 
payload I 
payload II 
data processing 
milestones 
calendar 
reference 
 

payload I

UV sensors 
DSLR camera 
mid size camera 
compact camera 
stereo 
line scan 
multi-spectral 
hyperspectral 
thermal imaging 
lidar scan 
 

Hyperspectral imaging system

Hyperspectral imaging is a very high spectral resolution form of multi-spectral imaging.

In hyperspectral imaging, one derives multiple images of the same area, through the effective use of multiple, contiguous, narrow band (for example, 10 nm Full Width Half Maximum = FWHM) optical filters.

A practical implementation of hyperspectral imaging is to be found in the Hyperion system developed by TRW.

 

 

 

 

 

 


 

In the TRW  system, a satellite orbits the earth and:

  • samples a 256 pixel line scan image, every 4.5 mS
  • uses an optical diffracting element, such as a diffraction grating, to derive a 242 bin optical spectral distribution from each pixel in the line scan, covering an optical spectral range from 400 nm to 2400 nm. Each bin = 10 nm FWHM spectral width.
  • uses 256 x 242 pixel CCD arrays, followed by a 12 bit analog-to-digital converter, to digitise the optically diffracted hyperspectral image information, for each line scan.

return to top

Implementation using commercially available imaging systems

The light for each of the three imaging systems is filtered using dichroic filters. Each pixel in the linescan is then split by wavelength, using a diffraction grating, into one of 256 bins. In order to increase the light level on each pixel, the dispersed images of successive line scans can be superimposed on the imaging CCD, with software then being used to extract the hyperspectral information from the superimposed data.

 

CCD visual

Alpha NIR

Merlin Thermal

 

wavelength

(H) x (V)

0.1 – 0.8

2,048 x 2,048

0.9 – 1.7

320 (H) x 256 (V)

1.5 – 5.0

320 (H) x 256 (V)

um

pixels x pixels

return to top

The TRW Hyperion system

The enhancement of the mineral rich regions can be clearly seen in the hyperspectral composite image, above right, relative to the true colour image, above left.

return to top


information derived from hyperion.pdf.

additional text © Barnard Microsystems Limited 2006 - 2008