Digital Image Processing Workflow

Atmospheric correction in the visible part of spectrum

• Cosmetic corrections

• Relative AC methods based on ground reflectance properties

• Absolute AC methods based on atmospheric process information

 Cosmetic corrections
• These procedures are not true AC techniques just to correct visible errors and
noise in image data.
• Cosmetic correction is achieved using especially designed filters and image
stretching and enhancement procedures
• Typical problems requiring cosmetic corrections are
qPeriodic line dropouts.
qLine stripping
qRandom noise or spike corrections
• Above effects can be identified visually and automatically

Periodic line dropouts

• Periodic line dropouts occur due to recording problems when one of the detectors of the sensor gives
wrong data or stop functioning.
• Landsat ETM has 16 detectors in all its bands. Expect thermal band.
• A loss of one of the detectors would result in every sixteenth scan line being a string of zeros results a
black line on the image.
How to restore….?
o First step to calculate the average DN value per san line for the entire scene
o Average DN value for each scan line is then compared with this scene average .
o Scan line deviating from the average by more than a designated threshold value is identified as defective.
o Replace the defective lines.
o For each pixel in a defective line an average DN is calculated using DNs for the corresponding pixel.
o The average DN is then substituted for the defective pixel.
o Resulting image is seems to be improved.
 Before correction

After correction
 Line Stripping
• Line stripping is far more common than line dropouts.
• Often occurs due to non identical detector response although the detector for all satellite sensors
are carefully calibrated and matched before the launch of satellite.
• With time response of some detectors may drift to higher or lower levels results every scan line
recorded by that detector is brighter or darker than other line.
• Its important to understand valid data are present in the defective line.
How to correct…..?
o Histogram matching.
o Separate histogram corresponding to each detector unit are constructed and matched.
o Taking one response as a standard and gain (Rate of increase of DN) and offset for all detector unit
are suitably adjusted and new DN values are compared and assigned.
o This yields a destriped image in which all DN values conform the reference level and scale.
 Random noise or spike noise
• The periodic line dropouts and striping are forms of non random noise that may
be recognized and restored by simple means.
• Random noise on the other hand requires a more sophisticated restoration
method such as digital filtering
• Such errors may be due to errors during transmission of data or temporary
disturbance.
• Some pixels acquire DN values that are much higher or lower than the
surrounding pixel.
• Can be detected mutually comparing neighbouring pixel value.
• If neighbouring values of pixel differ by more than a specific threshold margin it
is designated as a spike noise and DN is replaced by interpolation DN value based
on the surrounding pixel
•
 Relative AC methods
Two Reference Surfaces
• The output of this method is an image that matches a reflectance that is compatible with
the atmosphere of similar image taken on other previous date.
• This method works on the individual band/channel basis and valid to establish a uniform
comparison basis to study.
• For instance the evolution of non flux related parameters like indexes or land properties
derived directly or indirectly from a band.
• This method is relay on existence of at least one dark and one bright invariant area.
• Relative invariant areas are considered to retain their reflective properties over time .
• The atmospheric composition is unknown in the two imaged but their influence is
measurable by calculating difference in radiance between the two dates on the reflective
invariant areas.
• The procedure defines one image as master and other is slave.
Absolute AC methods based on Atmospheric processes
• These methods require a description of the components in the atmospheric
profile.
• The output of these methods is an image that matches the reflectance of
the ground pixels with maximum estimated error of 10%.if atmospheric
profiling is adequate enough.
• The advantage of these methods is that ground reflectance can be
evaluated under any atmospheric condition, altitude and relative geometry
between sun and satellite .
• Adopt the same procedure as discussed above.

Radiative transfer models

• Radiative transfer models can be used for computing either radiances or
irradiances for a wide variety of atmospheric and surface conditions.
• They require full description of the atmospheric components at fixed
altitudes.
• LOWTRAN , MODTRAN ,CODE 5S and 6S are all reference RTMs .Code 6S is the
most complete in the visible spectrum.
Data Input to RTMs
• Options that allow more accurate calculation of molecular absorption in
presence of multiple scattering
• The type of atmospheric path
• The kind of operating mode.
• Temperature and pressure profile.
• The water vapor zone, ozone , methane, nitrous oxide, carbon monoxide and
other gases altitude profile.
• Sensor type.
Band Transmission models for RS image processing
• RTMs process extensive AC calculations and are not normally used for image processing .
• Instead they are used on in individual cases.( study a specific target)
• The result of investigation can be used to build look up tables.
• The range of available options is very much focused on the problems routinely occurring
during image pre processing for example where the atmospheric description is unknown for
the image being processed.
• Material on earth have unique spectral signatures that are normally distorted by
atmospheric interactions along the path towards the satellite .
• If there is problem in the pixel information it should be corrected by the models used in
image processing .
• If such information did not provided by database then it should be corrected for the desired
data outputs.
• For example SMAC is a simplified version of code 6S . It is originally designed for NOAA
imagery and has been extended to include some high resolution satellite .
• But the data about ozone and aerosols is missing so it can be synthesized using sun
photometers which is operated in ground stations in many countries around the world and
produce the necessary information on an hourly basis.
 Standard Atmospheres
• Due to rapid dynamics of atmosphere in terms of temporal and spatial variation of its
constituents a processer of researcher need to define some common profiles corresponds to
average condition in different part of the world.
• Such fixed atmospheres is based on actual radiosoundings carried out at different research
sites resulting in so called standard atmospheres.
• RTMs have these standards built in to the system allowing the influence of different
constituents to be compared under strict simulations .
• For example the influence of water vapour in the thermal or aerosols in the visible spectrum.
• Which have a good correlation of the atmospheres around.
• Ozone is mainly concentrated between 15 to 30 km where most of the attenuation of the
ultraviolet take place.
• Water vapour is concentrated in the lower atmosphere in all cases, with a maximum close to
the earth surface.