Guillaume,
I understand you suggest that we fit a sensor model using GCPs and apply it
instead of using the methods I was suggesting.
I remember that we actually tried this method for an Hyperion image on the
Andes (lot of relief) and the result was very poor.
ref.
https://mail.google.com/mail/u/1/#search/label%3Aotb-users+alobolistas/140c98557c22c7c7

Instead, we applied the method described in Pala & Pons using another
software and the result was very good.
So I would say both methods are not equivalent, but I might be wrong. Note
that GCPs on wild areas (with no houses, roads etc)
are always scarce, perhaps not enough for a reasonable sensor model, but
sufficient for a simple 2nd order polynomial.
Agus



On Tue, Mar 21, 2017 at 10:02 AM, Guillaume Pasero <[email protected]>
wrote:

> Hi,
> On 03/21/2017 09:18 AM, Agustin Lobo wrote:
>
> I guess the second. 
> Examples:http://geog.uoregon.edu/amarcus/geog418/Labs/Lab06_rectification.htm
>
> The polynomial can be fit and applied using gdal_translate to
> incorporate the GCPs and then
> gdalwarp r near order 2
> to apply the transformation
>
> but the way of introducing the GCPs is kind of inconvenient.
>
> This problem is more about sensor model fitting. You have an image in
> sensor geometry and you try to ortho-rectify it on ground based on a set of
> GCPs. The OTB already has an application to assess a RPC sensor model based
> on GCPs.
>
> The polynomial model is a particular case of rational model, maybe setting 
> the denominator coefficients to [1,0,0,...,0] will do the trick.
>
> Guillaume
>
> Yes, I agree. But can this be done in practice? Also, is there a tool
> in OTB to actually fit such a rational polynomial with prescribed
> denominators?
>
> No, at the moment there is no way to estimate a polynomial transform based
> on GCPs. If the transform is already computed, it can be fed into a *.geom
> file to act as a RPC model with denominator coefficients set like proposed.
>
> Anyway, please consider also my suggestion of including the method described 
> in
> Pala, V., and X. Pons. 1995. “Incorporation of Relief in
> Polynomial-Based Geometric Correction.” Photogrametrics Engineering &
> Remote Sensing 61: 
> 935–44.https://www.researchgate.net/publication/244485309_Incorporation_of_relief_polynomial-based_geometric_corrections
>
> The sensor models in OTB (RPC based or physical) should already take into
> account elevation. The inverse localization function is modeled like :
>     (row,col) = F_inv(x,y,z)
>
>
> --
> <http://www.c-s.fr> *Guillaume PASERO*
> Responsable technique
> *Business Unit ESPACE & GeoInformation - Département Payload Data &
> Applications*
>
> *CS Systèmes d'Information*
> Parc de la Grande Plaine - 5, Rue Brindejonc des Moulinais - BP 15872
> 31506 Toulouse Cedex 05 - FRANCE
> +33 561 17 64 21 - [email protected]
>
> --
> --
> Check the OTB FAQ at
> http://www.orfeo-toolbox.org/FAQ.html
>
> You received this message because you are subscribed to the Google
> Groups "otb-users" group.
> To post to this group, send email to [email protected]
> To unsubscribe from this group, send email to
> [email protected]
> For more options, visit this group at
> http://groups.google.com/group/otb-users?hl=en
> ---
> You received this message because you are subscribed to the Google Groups
> "otb-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.
>

-- 
-- 
Check the OTB FAQ at
http://www.orfeo-toolbox.org/FAQ.html

You received this message because you are subscribed to the Google
Groups "otb-users" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/otb-users?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"otb-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to