Add tests with OpticalCalibration
- find a compatible remote image
- add test for parameters
- add test for pipelines
Activity
-
Newest first Oldest first
-
Show all activity Show comments only Show history only
- Cresson Remi added ci label
added ci label
- Cresson Remi added tests label
added tests label
- Author Owner
you can use this one
By Rémi Cresson on 2023-07-10T17:21:53 (imported from GitLab)
1 - Cresson Remi marked the checklist item find a compatible remote image as completed
marked the checklist item find a compatible remote image as completed
By Vincent Delbar on 2023-07-10T17:31:02 (imported from GitLab)
- Author Owner
@remicress may be you already noticed, but in
app.GetParameterValue()
for"acqui.*"
or"atmo.*"
we always get default values (same results with SPOT or Pléiades DIMAPs).
Even after exec and writing file !
Probably an OTB problem ? Nothing is returned byapp.GetParameters()
for these keys.
This is all in_auto_parameters
{'acqui.minute': 0, 'acqui.hour': 12, 'acqui.day': 1, 'acqui.month': 1, 'acqui.year': 2000, 'acqui.sun.elev': 90.0, 'acqui.sun.azim': 0.0, 'acqui.view.elev': 90.0, 'acqui.view.azim': 0.0, 'atmo.aerosol': 'noaersol', 'atmo.oz': 0.0, 'atmo.wa': 2.5, 'atmo.pressure': 1030.0, 'atmo.opt': 0.20000000298023224, 'atmo.radius': 2, 'atmo.pixsize': 1.0, 'level': 'toa', 'clamp': False, 'milli': False, 'in': <pyotb.Input object, from https://gitlab.orfeo-toolbox.org/orfeotoolbox/otb/-/raw/develop/Data/Baseline/OTB/Images/prTvOrthoRectification_pleiades-1_noDEM.tif>, 'out': '/tmp/test.tif'}
By Vincent Delbar on 2023-07-10T17:50:32 (imported from GitLab)
Edited by Cresson Remi Collapse replies - Author Owner
My bet was on the OSSIM removal but same results for DIMAP in OTB 7.4.1.
I wasn't able to create the input object with you link using this version, I guess vsicurl wasn't implemented yet.By Vincent Delbar on 2023-07-10T17:49:58 (imported from GitLab)
Edited by Cresson Remi - Author Owner
I think that
atmo.*
is used only whenlevel
istoc
.If that's the case, the application could have been implemented with
level.toc.atmo.*
instead👼 Now it is just a set of unused parameters, nothing is wrong with the logic in pyotb.
By Rémi Cresson on 2023-07-10T17:59:13 (imported from GitLab)
- Author Owner
Of course you're right, the atmo should not appear here, and this is due to this line of the
__sync_parameters
function.
May be we could an ugly special case here to avoid avoind toa + atmo params in summarize.But originally I was looking at the
acqui
params which are needed to compute TOA reflectance.
This looks like a dirty bug in the API.
We don't want this kind of trash in the summarize() outputBy Vincent Delbar on 2023-07-10T19:35:34 (imported from GitLab)
Edited by Cresson Remi - Author Owner
Result with
level="toc"
looks the same :{'milli': False, 'clamp': True, 'acqui.minute': 0, 'acqui.hour': 12, 'acqui.day': 1, 'acqui.month': 1, 'acqui.year': 2000, 'acqui.sun.elev': 90.0, 'acqui.sun.azim': 0.0, 'acqui.view.elev': 90.0, 'acqui.view.azim': 0.0, 'atmo.aerosol': 'noaersol', 'atmo.oz': 0.0, 'atmo.wa': 2.5, 'atmo.pressure': 1030.0, 'atmo.opt': 0.20000000298023224, 'atmo.radius': 2, 'atmo.pixsize': 1.0, 'level': 'toc', 'in': <pyotb.Input object, from /media/data/raster/airbus/phr/5976255101/IMG_PHR1B_MS_002/IMG_PHR1B_MS_202109290826464_SEN_5976255101-2_R1C1.TIF>}
By Vincent Delbar on 2023-07-10T19:37:38 (imported from GitLab)
Edited by Cresson Remi - Author Owner
So is the 6S computation always the same?
2023-07-10 21:42:02 (INFO) OpticalCalibration: Atmospheric correction parameters compute by 6S : AtmosphericCorrectionParameters (0x559094d46270) Atmospheric pressure : 1030 Water vapor amount : 2.5 Ozone amount : 0 Aerosol model : 0 Aerosol optical : 0.2
P.S. same when using Input obj or filepath.
By Vincent Delbar on 2023-07-10T19:48:42 (imported from GitLab)
Edited by Cresson Remi - Author Owner
So I don't know what to do with these atmo params.
By Vincent Delbar on 2023-07-11T12:28:00 (imported from GitLab)
- Author Owner
Nothing, just let them as they are. If they change in future OTB release (I may propose a MR one day), we will update the baseline at this point
By Rémi Cresson on 2023-07-11T14:59:58 (imported from GitLab)
- Cresson Remi added bug label
added bug label
- Author Owner
@remicress about the
acqui
parameters : this is lost when we use theInput
object, because of ExtractROI, looks like metadata are missing.
The values are correct when we pass a filepath directly to the OpticalCalibration app :{'milli': False, 'clamp': True, 'acqui.minute': 0, 'acqui.hour': 11, 'acqui.day': 15, 'acqui.month': 1, 'acqui.year': 2012, 'acqui.sun.elev': 23.836299896240234, 'acqui.sun.azim': 163.8679962158203, 'acqui.view.elev': 81.33570098876953, 'acqui.view.azim': 258.0060119628906, 'atmo.aerosol': 'noaersol', 'atmo.oz': 0.0, 'atmo.wa': 2.5, 'atmo.pressure': 1030.0, 'atmo.opt': 0.20000000298023224, 'atmo.radius': 2, 'atmo.pixsize': 0.0004999080556444824, 'level': 'toc', 'in': '/vsicurl/https://gitlab.orfeo-toolbox.org/orfeotoolbox/otb/-/raw/develop/Data/Baseline/OTB/Images/prTvOrthoRectification_pleiades-1_noDEM.tif'}
By Vincent Delbar on 2023-07-11T06:45:40 (imported from GitLab)
Edited by Cresson Remi Collapse replies - Author Owner
this bug has been fixed in OTB 8.1.2
By Rémi Cresson on 2023-07-11T06:45:40 (imported from GitLab)
- Author Owner
oooh yes I remember this issue now. thought it was merged a long ago ! but indeed the last release is kind of old.
By Vincent Delbar on 2023-07-11T12:27:18 (imported from GitLab)
- Cresson Remi created branch
107-add-tests-with-opticalcalibration
to address this issuecreated branch
107-add-tests-with-opticalcalibration
to address this issueBy Vincent Delbar on 2023-07-11T16:34:15 (imported from GitLab)
- Cresson Remi closed with commit 4e04b817
closed with commit 4e04b817