#6893 closed defect (fixed)
Stereographic projection not supported in ISIS3 write driver
Reported by: | swalter75 | Owned by: | warmerdam |
---|---|---|---|
Priority: | normal | Milestone: | 2.3.0 |
Component: | default | Version: | unspecified |
Severity: | normal | Keywords: | |
Cc: | thare@… |
Description
Please see proposed patch. Are there an reasons for it being left out?
Attachments (4)
Change History (15)
by , 7 years ago
Attachment: | gdal_isis3_stereographic.patch added |
---|
comment:1 by , 7 years ago
Trent, do you confirm that ISIS3 supports Stereographic ? Is there a reference for projections supported by ISIS3 ? Googling for "ISIS3 projections" brought me to http://jlaura.github.io/isis3/class_isis_1_1_projection.html which has a list of projections and I cannot see Stereographic in it
comment:2 by , 7 years ago
For whatever reason, ISIS3 was setup to only support "Polar Stereographic" not more generically just "Stereographic". I haven't looked at the code to see if there are differences in the equations but ISIS3 has a specific check such that the center lat cannot be 0. Thus mapping a "GDAL" Stereographic to an ISIS3 PolarStereographic should probably not be done. Keep this patch around but don't include yet. I will see if we can introduce "Stereographic" into ISIS3 but it will take some time to get it pushed in (if accepted).
resources: https://isis.astrogeology.usgs.gov/Object/Developer/group___map_projection.html polar St: https://isis.astrogeology.usgs.gov/Object/Developer/class_isis_1_1_polar_stereographic.html#details
comment:3 by , 7 years ago
Then we could at least check for "Stereographic" and abs(lat_0)=90 for the case that some software (like Vicar) does not always puts "polar_stereographic" in the label but only "stereographic" instead.
comment:4 by , 7 years ago
Or should we change the Vicar driver to check for the lat_0 and change the projection to "polar_stereographic"?
comment:5 by , 7 years ago
that is a good compromise - can you reissue the patch? I think I prefer the check "Stereographic" and abs(lat_0)=90.
Polar Stereo vs Stereo has always been a little confusing - thus updating the vicar driver might be good also.
by , 7 years ago
Attachment: | isis3_vicar_stereographic.patch added |
---|
Here is a new patch for ISIS3 and VICAR. It checks for Stereographic AND center_lat==90 and changes Proj. to Polar_Stereographic. I also optimized the NoData handling in VICAR a bit...
comment:7 by , 7 years ago
- Do we need in isis3_vicar_stereographic.patch the special case for EQUAL(pszProjection, SRS_PT_STEREOGRAPHIC) and abs(oSRS.GetNormProjParm(SRS_PP_LATITUDE_OF_ORIGIN, 0.0)) == 90 ? This is probably correct, but shouldn't such parameters already be mapped to SRS_PT_POLAR_STEREOGRAPHIC in GDAL ?
- Please use fabs() instead of abs() (holds for pds_stereographic.patch also)
- In vicardataset.cpp the value = poDS->GetKeyword("BLTYPE"); line doesn't seem to be used since afterwards there is if ( EQUAL(poDS->GetKeyword("BLTYPE"),"M94_HRSC") ) afterwards
comment:8 by , 7 years ago
If these cases are handled everywhere in GDAL we can omit changing the ISIS3 driver (I tried a quick test with gdalwarp and it also automatically changes "Stereographic" to "Polar_Stereographic" at poles).
Please see my new patch for the other two points (now only VICAR and PDS drivers are changed).
by , 7 years ago
Attachment: | vicar_pds_stereographic.patch added |
---|
This one only changes the VICAR and PDS drivers to change the Stereographic projection to Polar_Stereographic if (lat_0 = +-90)
comment:10 by , 7 years ago
Milestone: | → 2.3.0 |
---|
path for adding stereographic support