SREF X STDSREF Parameters arcsi.py

[Felipe Ramos Dantas]

Dear,

when adding the STDSREF parameter in the atmospheric correction for L8 or S2 images I noticed that the resulting image gets a kind of light as shown in the attached figure (Layer problem) which does not happen if I use the SREF parameter. I am not sure if this is a problem, but neighborly it makes the image very strange. Could someone explain to me how this works? If this result is normal?

Sincerely !

[Peter Bunting]

Dear Felipe,

I’ve not seen that effect over the flat areas before. The algorithm being applied here is an implementation of the following paper - basically a topographic correction:

Shepherd, J., Dymond, J. (2003). Correcting satellite imagery for the variance of reflectance and illumination with topography International Journal Of Remote Sensing 24(17), 3503 3514. https://dx.doi.org/10.1080/01431160210154029

In terms of what could be causing this, I suggest the following things:

1. Check the DEM you are using - what resolution is it? It must have texture over the flat areas.
2. What is the sun angle? I am guessing the sun angle is low as that is where topographic correction algorithms struggle and in this case over corrects the steep slopes and highlighting texture within the ‘flat’ areas.

Best wishes,

Pete

> --
> You received this message because you are subscribed to the Google Groups "RSGISLib Support" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to rsgislib-suppo...@googlegroups.com.
> To view this discussion on the web, visit https://groups.google.com/d/msgid/rsgislib-support/df6d0bbf-ba97-4992-9cfa-1fde7d35cbdf%40googlegroups.com.
> <Layer problem.jpg><layer no problem.png>

[Felipe Ramos Dantas]

Hi Peter Thanks for the feedback. I was using an ASTER V3 DEM when the above image problems occurred. I did the same procedure with the AW3D30 (ALOS) DEM and sentinel image and the problem did not occur. Thanks again for the help.Hi Peter Thanks for the feedback. I was using an ASTER V3 DEM when the above image problems occurred. I did the same procedure with the AW3D30 (ALOS) DEM and sentinel image and the problem did not occur. Thanks again for the help.Hi Peter Thanks for the feedback. I was using an ASTER V3 DEM when the above image problems occurred. I did the same procedure with the AW3D30 (ALOS) DEM and sentinel image and the problem did not occur. Thanks again for the help.
Hi Peter,

Thanks for the feedback. I replaced DEM ASTER V3 (problem generator) with ALOS (AW3D30) and the problem was fixed.

Thanks again for the help!

Em quarta-feira, 13 de novembro de 2019 18:40:08 UTC-3, Peter Bunting escreveu:

Dear Felipe,

I’ve not seen that effect over the flat areas before. The algorithm being applied here is an implementation of the following paper - basically a topographic correction:

Shepherd, J., Dymond, J. (2003). Correcting satellite imagery for the variance of reflectance and illumination with topography International Journal Of Remote Sensing 24(17), 3503 3514. https://dx.doi.org/10.1080/01431160210154029

In terms of what could be causing this, I suggest the following things:

1. Check the DEM you are using - what resolution is it? It must have texture over the flat areas.
2. What is the sun angle? I am guessing the sun angle is low as that is where topographic correction algorithms struggle and in this case over corrects the steep slopes and highlighting texture within the ‘flat’ areas.

Best wishes,

Pete


> On 13 Nov 2019, at 19:28, Felipe Ramos Dantas <felipe...@ifpi.edu.br> wrote:
>
> Dear,
>
> when adding the STDSREF parameter in the atmospheric correction for L8 or S2 images I noticed that the resulting image gets a kind of light as shown in the attached figure (Layer problem) which does not happen if I use the SREF parameter. I am not sure if this is a problem, but neighborly it makes the image very strange. Could someone explain to me how this works? If this result is normal?
>
> Sincerely !
>
> --
> You received this message because you are subscribed to the Google Groups "RSGISLib Support" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to rsgislib...@googlegroups.com.