Abstract
High-resolution satellite images must undergo a geometric rectification process in order to be used for metrical purposes. This operation, called orthorectification, is necessary because of deformations mainly due to camera distortions and acquisition geometry. To correctly orthorectify an image, it is necessary to accurately reconstruct the photogrammetric-acquisition characteristics and the image position with respect to a reference system connected to the ground. This operation, called orientation, can be done using various mathematical models such as rigorous, rational polynomial function (RPF), and rational polynomial coefficient, or, according to some authors, rapid positioning coefficient (RPC) models. Orientation and orthorectification are usually performed within specific commercial software, but in QGIS, these complex operations can be performed using the open libraries of the Orfeo Tool Box (OTB). Unfortunately, instructions given by OTB developers lead to scarce results. In fact, the procedure proposed in OTB does not allow for the full exploitation of the potential of RPC models, on which OTB itself is based. As OTB is open-source software, a plugin was developed to overcome these limitations and exploit its full potential. In fact, OTB interfaces are unfortunately essential, and some necessary functions are missing. Therefore, a new QGIS plugin was developed in order to run the entire process in the most photogrammetrically and statistically correct way, and, at the same time, to simplify the relative procedures.
Highlights
In the last twenty years, the availability of high- and very-high-resolution satellite images has stimulated methodological research for accurate geometric correction and the orientation of the images themselves
We studied the possibilities of geometric correction made available by the Orfeo Tool Box (OTB), a library of open-source algorithms implemented within the GIS software QGIS, and we verified how some of the initial parameters of the rapid positioning coefficient (RPC) model could be improved to obtain results equal to or better than those of commercial software
OTB does not currently have a function for the collimation and management of ground control point (GCP) and check points (CPs), so in this experiment, the GCPs were collimated in the QGIS georeferencer, but image coordinates were not expressed according to OTB convention
Summary
In the last twenty years, the availability of high- and very-high-resolution satellite images has stimulated methodological research for accurate geometric correction and the orientation of the images themselves. We studied the possibilities of geometric correction made available by the Orfeo Tool Box (OTB), a library of open-source algorithms implemented within the GIS software QGIS, and we verified how some of the initial parameters of the RPC model could be improved to obtain results equal to or better than those of commercial software. This approach was implemented in a new QGIS plugin that was developed by the authors and that allows for the entire process to be carried out, from data input to the estimation of the actual accuracy of the results, which is presently not possible. This paper describes the improvements that we propose to the OTB orthorectification process, how they were implemented in our plugin, and the first obtained results, comparing them with those that could be obtained under the same condition using commercial software that is considered as a reference by the academic world
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
