A 3001*3022 IKONOS 3 bands BGR image of Heron island, GBR, Australia
4SM licensekey is 522 Euros for this image
Please refer to "Multiscale evaluation of IKONOS data for classification of tropical coral reef environmennts"
by Andrefouet et al. Remote Sensing of Environment, 2003
Wavelengths used here are 480, 565, 640 nm         Ikonos wavelengths     The NIR band is not available
An exceptional image
home
Work done in 2005


 
1 - NO NEED for field data, nor for atmospheric correction
2 - this is demonstrated in this website, using a variety of hyper/multi spectral data
 
Requirements are
1 - homogeneous water body and atmosphere
2 - some coverage of optically deep water
3 - some coverage of dry land
 
Problems are
1 - the precision on estimated depth is found wanting, because the noise-equivalent change in radiance  of accessible data is too high for shallow water column correction work 
2 - radiance data should be preprocessed by the provider at level 1 in order to improve S/N ratio
3 - exponential decay: the deeper/darker the bottom, the poorer the performances
 
So
I keep digging
until suitable data
become available
 
This image was copied to me already-deglinted by USF:
very bad deglinting ==>still very much needs deglinting.
I wish I had access to the original raw 4-bands image instead of this radiometric tuttifruti.
I lack seatruth depths in order to ascertain operational wavelengths
for Ikonos wavebands in 4SM.
Therefore, although the optical calibration seems quite straightforward,
depth results might very well need to be multiplied by a significant CoefZ correction factor.

 


Raw image BGR TCC
histeq enhancement


Deglinted image BGR TCC
histeq enhancement


As it reached us,
this image was already "deglinted"
  • by some unknown process which only removed the swell fabric of the glint
  • this is why(?) the NIR band is not available
  • but a lot of glint remains in the form of very high level of noise in all three bands 

 This noise is still fairly well correlated
among all 3 bands
  • therefore, it can be removed to a certain extent by 4SM
  • this proved to be quite challenging, but efficient
  • it involved quite some in-depth improvements of the deglinting process 


image Z stepped


image Z


image B


image LBS normalized


image LBS: RGB true color composite


image LBS chequered

   

 image WZ
RED where 3-bands modeling was possible
GREEN where onl Green/Blue modeling was possible

 image WR
illustrates the diameter, from 3 to 11,
of the circular smart-smoothing window
  PROFILE ACROSS THE CHANNEL: 
Z and spectral LBS
across the reef rim, through the channel,
then across the next reef rim 

 

SEAWEED_1: Z and spectral LBS

green bottoms around pixel 1265 and pixel 1450
in a blue bottom dominated area of the lagoon
topography is ~level at ~2 m

SEAWEED_2: Z and spectral LBS

green bottoms from pixel 1300 to pixel 1540
in a blue bottom dominated area of the lagoon
topography is ~level at ~2.5 m

PINNACLES 1: Z and spectral LBS

left half is blue bottoms dominated
right half is green bottoms dominated 

 
PINNACLES 2: Z and spectral LBS

right half is blue bottoms dominated
left half is green bottoms dominated 

 
 
Optical calibration

Oceanic IB water type of Jerlov: 
very clear waters
fairly good calibration in spite of 
  • poor radiometric quality
  • NIR band is not available
  • mind bogging "deglinting"quasi-absence of bare land pixels 
 
Operational wavelengths
 for Ikonos visible wavebands at Heron Island
are tentatively set to 
WLblue=  495.2 nm
WLgreen=575.9 nm
WLred=    667.4 nm
WLnir=    780.0 nm