Images vs DEMs in GeoTiff

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

Images vs DEMs in GeoTiff

Gail-2

Hello,

 

Is there currently anyway to determine if a given GeoTIFF file contains pixels or posts?

In other words, is there a tag that states that the file is an image or a DEM? If not a specific tag, then perhaps a combination of tags?

 

Thank you,

Gail


_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff
Reply | Threaded
Open this post in threaded view
|

Re: Images vs DEMs in GeoTiff

Frank Warmerdam
Nagle, Gail A (US SSA) wrote:
> Hello,
>
>  
>
> Is there currently anyway to determine if a given GeoTIFF file contains
> pixels or posts?
>
> In other words, is there a tag that states that the file is an image or
> a DEM? If not a specific tag, then perhaps a combination of tags?

Gail,

Yes, the GTRasterTypeGeoKey geokey can have the value RasterPixelIsArea or
RasterIsPoint.

Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, [hidden email]
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo, http://osgeo.org

_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff
Reply | Threaded
Open this post in threaded view
|

RE: Images vs DEMs in GeoTiff

Gail-2
In reply to this post by Gail-2
Hi Cody,

Thanks for adding this info. If the misinterpretation of the value of the GTRasterTypeGeoKey has become commonplace, I do think a better solution is available in the 1.5 Draft spec of GeoTIFF. We have a copy of a document numbered Specification version 1.9.3 dated 14 June 2001 from Niles Ritter. It contains the following proposed key in section 4.4.1:

DataInterpretationGeoKey
Key ID = 1103
Type = SHORT (code)
Values:
DataMultispectral=1
DataElevation=2
DataVector=3
DataThematic=4

I don't know if the 1.5 Draft is still being considered or not. But it does seem that the power of GeoTIFF is such that it is being used in ways that were not originally planned for. So having a definitive key describing the content of the file would be very useful.

In the meantime, when we are trying to determine what kind of data we have, an image or a DEM, we are going to look for the value RasterPixelIsPoint and assume it is a DEM and try to load it as such. Likewise, if the value is RasterPixelIsArea, we will try to load it as an image. If the user is not happy with our interpretation of the data, we will provide a way for the user to specify how to load the data: DEM or Image.

Thanks for taking the time to weigh in on this. Things do get complicated!

-Gail


-----Original Message-----
From: Cody Benkelman [mailto:[hidden email]]
Sent: Wednesday, June 25, 2008 6:58 PM
To: Nagle, Gail A (US SSA)
Subject: FW: [Geotiff] Images vs DEMs in GeoTiff

Gail

I never like to disagree with Frank Warmerdam, but I am very confident there
are GeoTIFF images that are examples of raster images (not DEM data) out
there in the world with the GTRasterTypeGeoKey geokey set to values of both
RasterPixelIsArea and RasterIsPoint (not "both in one file" but "examples of
both").

That may not be what he and others *intended* for usage of that field, but I
believe it has (in many cases) been interpreted as "is the precise
coordinate of the pixel in the CENTER of the pixel or at the CORNER of the
pixel?".

Unfortunately this is a tangent and doesn't answer your question - I do not
know if there is another tag that WILL specify if a GeoTIFF is a DEM (if so,
Frank must have misunderstood your question, or he would have offered it),
but I didn't want you to go too far with his response.

Having said all that, I *may* be wrong so please verify with others, but I'm
concerned that if you assume "GTRasterTypeGeoKey = RasterIsPoint" implies
"data = DEM" you'll run into problems.

Best wishes

Cody A. Benkelman
ASPRS Certified Mapping Scientist - Remote Sensing (#RS-144)
17 Willowbrook Close
Whitefish, Montana  59937
cell (406) 270-1176



-----Original Message-----
From: [hidden email]
[mailto:[hidden email]] On Behalf Of Frank Warmerdam
Sent: Wednesday, June 25, 2008 5:36 PM
To: Nagle, Gail A (US SSA)
Cc: [hidden email]
Subject: Re: [Geotiff] Images vs DEMs in GeoTiff

Nagle, Gail A (US SSA) wrote:
> Hello,
>
>  
>
> Is there currently anyway to determine if a given GeoTIFF file
> contains pixels or posts?
>
> In other words, is there a tag that states that the file is an image
> or a DEM? If not a specific tag, then perhaps a combination of tags?

Gail,

Yes, the GTRasterTypeGeoKey geokey can have the value RasterPixelIsArea or
RasterIsPoint.

Best regards,
--
---------------------------------------+--------------------------------
---------------------------------------+------
I set the clouds in motion - turn up   | Frank Warmerdam,
[hidden email]
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo, http://osgeo.org

_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff



_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff
Reply | Threaded
Open this post in threaded view
|

Re: Images vs DEMs in GeoTiff

Frank Warmerdam
Cody / Gail,

Cody is right.  I really only read your question as far as "contains pixels or
posts" before I answered.  The GTRasterTypeGeoKey cannot be reliably used
to differentiate image data from DEM data.

Note there are many types of raster data beyond "visual image", and DEM.

Best regards,

Nagle, Gail A (US SSA) wrote:

> Hi Cody,
>
> Thanks for adding this info. If the misinterpretation of the value of the GTRasterTypeGeoKey has become commonplace, I do think a better solution is available in the 1.5 Draft spec of GeoTIFF. We have a copy of a document numbered Specification version 1.9.3 dated 14 June 2001 from Niles Ritter. It contains the following proposed key in section 4.4.1:
>
> DataInterpretationGeoKey
> Key ID = 1103
> Type = SHORT (code)
> Values:
> DataMultispectral=1
> DataElevation=2
> DataVector=3
> DataThematic=4
>
> I don't know if the 1.5 Draft is still being considered or not. But it does seem that the power of GeoTIFF is such that it is being used in ways that were not originally planned for. So having a definitive key describing the content of the file would be very useful.
>
> In the meantime, when we are trying to determine what kind of data we have, an image or a DEM, we are going to look for the value RasterPixelIsPoint and assume it is a DEM and try to load it as such. Likewise, if the value is RasterPixelIsArea, we will try to load it as an image. If the user is not happy with our interpretation of the data, we will provide a way for the user to specify how to load the data: DEM or Image.
>
> Thanks for taking the time to weigh in on this. Things do get complicated!
>
> -Gail
>
>
> -----Original Message-----
> From: Cody Benkelman [mailto:[hidden email]]
> Sent: Wednesday, June 25, 2008 6:58 PM
> To: Nagle, Gail A (US SSA)
> Subject: FW: [Geotiff] Images vs DEMs in GeoTiff
>
> Gail
>
> I never like to disagree with Frank Warmerdam, but I am very confident there
> are GeoTIFF images that are examples of raster images (not DEM data) out
> there in the world with the GTRasterTypeGeoKey geokey set to values of both
> RasterPixelIsArea and RasterIsPoint (not "both in one file" but "examples of
> both").
>
> That may not be what he and others *intended* for usage of that field, but I
> believe it has (in many cases) been interpreted as "is the precise
> coordinate of the pixel in the CENTER of the pixel or at the CORNER of the
> pixel?".
>
> Unfortunately this is a tangent and doesn't answer your question - I do not
> know if there is another tag that WILL specify if a GeoTIFF is a DEM (if so,
> Frank must have misunderstood your question, or he would have offered it),
> but I didn't want you to go too far with his response.
>
> Having said all that, I *may* be wrong so please verify with others, but I'm
> concerned that if you assume "GTRasterTypeGeoKey = RasterIsPoint" implies
> "data = DEM" you'll run into problems.



--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, [hidden email]
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo, http://osgeo.org

_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff
Reply | Threaded
Open this post in threaded view
|

RE: Images vs DEMs in GeoTiff

Gail-2
In reply to this post by Gail-2
Frank,

Thank you for the clarification. Given the fact that there are many
types of raster data and the increasing use of automation to process
data, are there any plans to revise the GeoTiff standard to allow the
creator to specify how to interpret the data stored in a GeoTIFF file?
At the moment, a reader, computer or human, must guess.

Thanks,
-Gail


-----Original Message-----
From: Frank Warmerdam [mailto:[hidden email]]
Sent: Wednesday, June 25, 2008 8:18 PM
To: Nagle, Gail A (US SSA)
Cc: [hidden email]; [hidden email]
Subject: Re: [Geotiff] Images vs DEMs in GeoTiff

Cody / Gail,

Cody is right.  I really only read your question as far as "contains
pixels or
posts" before I answered.  The GTRasterTypeGeoKey cannot be reliably
used
to differentiate image data from DEM data.

Note there are many types of raster data beyond "visual image", and DEM.

Best regards,

Nagle, Gail A (US SSA) wrote:
> Hi Cody,
>
> Thanks for adding this info. If the misinterpretation of the value of
the GTRasterTypeGeoKey has become commonplace, I do think a better
solution is available in the 1.5 Draft spec of GeoTIFF. We have a copy
of a document numbered Specification version 1.9.3 dated 14 June 2001
from Niles Ritter. It contains the following proposed key in section
4.4.1:

>
> DataInterpretationGeoKey
> Key ID = 1103
> Type = SHORT (code)
> Values:
> DataMultispectral=1
> DataElevation=2
> DataVector=3
> DataThematic=4
>
> I don't know if the 1.5 Draft is still being considered or not. But it
does seem that the power of GeoTIFF is such that it is being used in
ways that were not originally planned for. So having a definitive key
describing the content of the file would be very useful.
>
> In the meantime, when we are trying to determine what kind of data we
have, an image or a DEM, we are going to look for the value
RasterPixelIsPoint and assume it is a DEM and try to load it as such.
Likewise, if the value is RasterPixelIsArea, we will try to load it as
an image. If the user is not happy with our interpretation of the data,
we will provide a way for the user to specify how to load the data: DEM
or Image.
>
> Thanks for taking the time to weigh in on this. Things do get
complicated!

>
> -Gail
>
>
> -----Original Message-----
> From: Cody Benkelman [mailto:[hidden email]]
> Sent: Wednesday, June 25, 2008 6:58 PM
> To: Nagle, Gail A (US SSA)
> Subject: FW: [Geotiff] Images vs DEMs in GeoTiff
>
> Gail
>
> I never like to disagree with Frank Warmerdam, but I am very confident
there
> are GeoTIFF images that are examples of raster images (not DEM data)
out
> there in the world with the GTRasterTypeGeoKey geokey set to values of
both
> RasterPixelIsArea and RasterIsPoint (not "both in one file" but
"examples of
> both").
>
> That may not be what he and others *intended* for usage of that field,
but I
> believe it has (in many cases) been interpreted as "is the precise
> coordinate of the pixel in the CENTER of the pixel or at the CORNER of
the
> pixel?".
>
> Unfortunately this is a tangent and doesn't answer your question - I
do not
> know if there is another tag that WILL specify if a GeoTIFF is a DEM
(if so,
> Frank must have misunderstood your question, or he would have offered
it),
> but I didn't want you to go too far with his response.
>
> Having said all that, I *may* be wrong so please verify with others,
but I'm
> concerned that if you assume "GTRasterTypeGeoKey = RasterIsPoint"
implies
> "data = DEM" you'll run into problems.



--
---------------------------------------+--------------------------------
------
I set the clouds in motion - turn up   | Frank Warmerdam,
[hidden email]
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo,
http://osgeo.org


_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff
Reply | Threaded
Open this post in threaded view
|

Re: Images vs DEMs in GeoTiff

Frank Warmerdam
Nagle, Gail A (US SSA) wrote:
> Frank,
>
> Thank you for the clarification. Given the fact that there are many
> types of raster data and the increasing use of automation to process
> data, are there any plans to revise the GeoTiff standard to allow the
> creator to specify how to interpret the data stored in a GeoTIFF file?
> At the moment, a reader, computer or human, must guess.

Gail,

I am not aware of any efforts to do this for GeoTIFF.  There are general
metadata standards that attempt to accomplish this sort of thing, and they
might be applied to GeoTIFFs to, but I don't generally have much faith in
this sort of thing.

Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, [hidden email]
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo, http://osgeo.org

_______________________________________________
Geotiff mailing list
[hidden email]
http://lists.maptools.org/mailman/listinfo/geotiff