Automated Registration and Orthorectification Package
Note: this is a software that was developed at USDA (United States Department of Agriculture). I'm just hosting the source to make it easier for others to consume it directly. I'm not the author of this software.
Here is the link for the original entry
Accurate geo-referencing information is a basic requirement for combining remote satellite imagery with other geographic information. To detect changes in time-series satellite images, it is extremely important for the images to be precisely co-registered and orthorectified, so that images acquired from different sensors and dates can be compared directly.
Precise registration relates satellite images to the ground reference based on carefully selected ground control points between the image and corresponding ground objects. Co-registration matches two images based on the tie points in the images. Although most commercial remote sensing software packages provide precise registration or co-registration, the selection of control points or tie points needs to be done manually. This is a time-consuming process that is not feasible for a large amount of data processing. Even though some satellite data are precisely registered and orthorectified, mismatches of up to a few pixels are common in early data sources.
The topographical variations of the earth’s surface and the satellite view zenith angle affect the pixel’s distance projected onto the satellite image. The distortion inherent in the image is determined by topographical elevation. For example, an off-nadir view of 5 degrees at a site with a terrain height of 1000 meters can cause displacement of 120 meters (about 4 Landsat Thematic Mapper (TM) pixels). The orthorectification process is used to correct the pixel displacement caused by the topographical variations at the off-nadir viewing and to make the image orthographic, with every pixel in its correct location regardless of elevation and viewing direction.
We developed an automated registration and orthorectification package (AROP) for Landsat and Landsat-like data processing. It uses precisely registered and orthorectified Landsat data (e.g., GeoCover or recently released free Landsat Level 1T data from the USGS EROS data center) as the base image to co-register and orthorectify the warp image and thus to make geo-referenced time-series images consistent in the geographic extent, spatial resolution, and projection. This package has been tested on the Landsat Multi-spectral Scanner (MSS), TM and Enhanced TM Plus (ETM+), Terra ASTER, CBERS CCD, and IRS-P6 AWiFS data.
The development of the AROP package was supported by the U.S. Geological Survey (USGS) Landsat Science Team project and the NASA EOS project. The package was initially developed at the NASA Goddard Space Flight Center (from September 2005 to June 2011). Previous version can be downloaded from the NASA GSFC Landsat Ecosystem Disturbance Adaptive Processing System (LEDAPS) web site. Further improvement and continuous maintenance are now being undertaken in the Hydrology and Remote Sensing Laboratory, Agricultural Research Service, U.S. Department of Agriculture (USDA) by Dr. Feng Gao.
Overview
Accurate geo-referencing information is a basic requirement for combining remote satellite imagery with other geographic information. To detect changes in time-series satellite images, it is extremely important for the images to be precisely co-registered and orthorectified, so that images acquired from different sensors and dates can be compared directly. Precise registration relates satellite images to the ground reference based on carefully selected ground control points between the image and corresponding ground objects. Co-registration matches two images based on the tie points in the images. Although most commercial remote sensing software packages provide precise registration or co-registration, the selection of control points or tie points needs to be done manually. This is a time-consuming process that is not feasible for a large amount of data processing. Even though some satellite data are precisely registered and orthorectified, mismatches of up to a few pixels are common in early data sources. The topographical variations of the earth’s surface and the satellite view zenith angle affect the pixel’s distance projected onto the satellite image. The distortion inherent in the image is determined by topographical elevation. For example, an off-nadir view of 5 degrees at a site with a terrain height of 1000 meters can cause displacement of 120 meters (about 4 Landsat Thematic Mapper (TM) pixels). The orthorectification process is used to correct the pixel displacement caused by the topographical variations at the off-nadir viewing and to make the image orthographic, with every pixel in its correct location regardless of elevation and viewing direction. We developed an automated registration and orthorectification package (AROP) for Landsat and Landsat-like data processing. It uses precisely registered and orthorectified Landsat data (e.g., GeoCover or recently released free Landsat Level 1T data from the USGS EROS data center) as the base image to co-register and orthorectify the warp image and thus to make geo-referenced time-series images consistent in the geographic extent, spatial resolution, and projection. This package has been tested on the Landsat Multi-spectral Scanner (MSS), TM and Enhanced TM Plus (ETM+), Terra ASTER, CBERS CCD, and IRS-P6 AWiFS data. The development of the AROP package was supported by the U.S. Geological Survey (USGS) Landsat Science Team project and the NASA EOS project. The package was initially developed at the NASA Goddard Space Flight Center (from September 2005 to June 2011). Previous version can be downloaded from the NASA GSFC Landsat Ecosystem Disturbance Adaptive Processing System (LEDAPS) web site. Further improvement and continuous maintenance are now being undertaken in the Hydrology and Remote Sensing Laboratory, Agricultural Research Service, U.S. Department of Agriculture (USDA) by Dr. Feng Gao.
Installation
(tested in Linux bash)
gzip -d public_AROP_v2.2.6.tar.gz tar -xvf public_AROP_v2.2.6.tar cd public_AROP_v2.2.6 source env.sh make make install
You will see an executable program named "ortho" under sub-directory "public_AROP_v2.2.6/bin" if programs were compiled successfully.
Usage
Usage: ortho [-r|-o|-b|-v] -r do registration only -o do orthorectification only (assume warp image has been registrated) -b do both registration and othorectification (recommend for all L1G data) -v verify image matching
Input parameter file defines the detailed information for base, warp and output image. The base image and the matching band from warp images will be used for the area-based tie point searching base on the tie point control file. Please check appendix A.1 and A.2 published in the Journal of Applied Remote Sensing by Gao et al. (2009) for descriptions of input parameter file and tie point control file. The updated input parameter file is provided in readme file under “doc” subdirectory.
Reference
Gao., F., Masek, J. G., Wolfe, R. F., An automated registration and orthorectification package for Landsat and Landsat-like data processing, Journal of Applied Remote Sensing, Vol.3, 033515, doi: 10.1117/1.3104620, 2009. (copyright SPIE JARS)
Unconditional Release
The package is now being released at no cost to the public for further development. ARS is releasing the AROP so that interested parties can use the software tool for their own needs and purposes. ARS does not foresee providing monetary or technical support to refine, adapt, or use this software tool, and provides no warranty for its use for any purpose. ARS does not reserve any rights or interests in the work that may be performed by others to refine or adapt it. ARS does reserve the right to continue its own refinement of the current version of the software package at a later date, should program needs require it.