ATCOR 4 - Features
ATCOR 4 is a software for the atmospheric/topographic correction of airborne scanner data. This may be used in four different geometric situations, where the (optional) geo- rectification is to be done by PARGE or a compatible software:
- 1. flat terrain, no geocoding (raw geometry)
- 2. flat terrain, geocoded/rectified image on flat surfaces (one altitude)
- 3. rugged terrain, DEM and related layers required, ortho-rectified image
- 4. rugged terrain, geocoded (non resampled), DEM and related layers provided in raw geometry
- Bottom of atmosphere reflectance calculation with constant atmospheric conditions or spatially varying aerosol conditions.
- Retrieval of atmospheric water vapor column for sensors with water vapor bands (around 820/940/1130 nm), using the APDA method.
- Aerosol optical thickness retrieval using the dark dense vegetation (DDV) or the shadow based (SHAOT) method.
- Statistical haze removal through use of a fully automatic algorithm that masks haze and cloud regions and removes haze of land areas.
- Cirrus cloud removal (requires a narrow 1.38 micron band).
- De-shadowing of cloud/building cast shadow areas using image-based shade detection routines.
- Empirical BRDF correction of irradiance effects.
- Model based observation angle dependent BRDF correction (BREFCOR method)
- Optional quick topographic correction (without atmospheric correction).
- Inflight radiometric calibration to calculate gain and offset from known target surface reflectances.
- Inflight at-sensor radiance validation using ground target reflectances.
- Automatic classification of spectral surface reflectance (program SPECL) using 12 surface cover templates.
- Surface emissivity and surface (brightness) temperature maps for thermal band sensors.
- Spectral polishing of reflectance cube (only hyperspectral imagery).
- Value added products side outputs (LAI, FAPAR, etc.).
- Correction of spectral "smile" (across track variation of the spectral position of spectral bands for hyperspectral instruments),
- Adaption of solar reference spectrum,
- Diffuse and direct illumination components side outputs,
- The atmospheric database includes a wide range of pre-calculated radiative transfer runs for different weather conditions and sun angles employing the Modtran®-5 code.
- Works for hyperspectral, multispectral, thermal, and panchromatic imagery.
- The atmospheric database (look-up-tables of radiative transfer calculations with Modtran®-5) covers a wide range of weather conditions and sun angles, calculated using the DISORT 8-stream method for multiple scattering. (>>See further details)
All outputs are stored as band sequential binary data with standard ENVI header, i.e.:
- Surface reflectance cube
- Surface (brightness) temperature and emissivity (if multispectral thermal bands)
- Visibility index map (corresponds to total optical depth at 550 nm) and aerosol optical thickness map
- Water vapor map (if required water vapor channels are available, e.g. at 940 nm)
- Spectral reflectance / emissivity surface cover map (10-12 classes)
- Value added channels: SAVI, LAI, FPAR, albedo, radiation and heat fluxes
- Spectral direct and diffuse irradiance cubes
- Masks of haze/cloud/water, and saturated pixels
- AISA, APEX, AVIRIS, CASI, Hymap, Hyspex, Headwall Hyperspec, Resonon Pika, Daedalus, Leica ADS, ...
Any user-specified sensor can easily be included by copying one of the existing "sensor*.dat" files and modifying it appropriately. Tools exist to generate the spectral channel filter functions for Gaussian or square wave response curves.
(Note: tilted sensors are not supported in the atmospheric database)Technical Conditions:
- Unix / Linux / Windows / MacOSX
- IDL Version 8.4 or higher or the free IDL Virtual Machine (provided with the software)
- PARGE Version 3.x or compatible solution if orthorectified imagery is required.
- Large amount of RAM and a 64bit system are highly recommended in interactive mode (tiling capability to reduce memory requirements is available).
- maximum RAM requirements: 18*l*c [MB] (flat terrain) and 35*l*c [MB] (rugged terrain) allocated to IDL,
where l=lines, c=columns of image.
Example rugged terrain: l=3000, c=2000 requires 210 MB RAM
- 6 GB of disk space for atmospheric database
- ENVI license recommended, but not required
Example flat terrain: l=3000, c=2000 requires 108 MB RAM
- Support includes updates, bug fixes and e-mail / phone support for fully licensed users. Details about the software methodology can be found in the User Manual and in various published papers.