[gdal-dev] hdf5 data to geotiff(the coordinates)

maven apache apachemaven0 at gmail.com
Mon Jan 11 23:00:51 EST 2010


2010/1/12 Frank Warmerdam <warmerdam at pobox.com>

> maven apache wrote:
>
>>
>>
>> 2010/1/12 Frank Warmerdam <warmerdam at pobox.com <mailto:
>> warmerdam at pobox.com>>
>>
>>
>>    maven apache wrote:
>>
>>           It suggests that GDAL is not recognising the georeferencing
>>        of the
>>           data.  The actual pixel values will hopefully still be accessed
>>           properly.
>>
>>           I would actually need to see the whole gdalinfo report on the
>>        subdataset
>>           to be certain that GDAL is not finding another form of
>>        georeferencing,
>>           like GCPs or geolocation arrays.  These will not be reflected
>>        in the
>>           corner coordinates report (unfortunately).
>>
>>        Thank you very much , I will paste the result.
>>        The whole gdalinfo metadata can be found here:
>>        http://dpaste.com/143948/
>>        And the info of  SUBDATASET_4_NAME  can be found here:
>>        http://dpaste.com/143949/
>>
>>
>>    "Maven",
>>
>>    It seems that GDAL does not recognise the metadata on this file that
>>    contains the georeferencing.  However, you can fairly easily apply the
>>    georeferencing as part of the translation something like this:
>>
>>    gdal_translate \
>>
>> HDF5:"d23.he5"://HDFEOS/GRIDS/ColumnAmountO3/Data_Fields/ColumnAmountO3
>>    \
>>     -a_ullr -180 90 180 -90
>>
>> Oh thanks, it works!
>> But I have two questions:
>> 1) how do you know the -a_ullr parameter?
>>
>
> Maven,
>
> I read through the metadata reported by gdalinfo:
>
>  ColumnAmountO3:GCTPProjectionCode=0
>  ColumnAmountO3:Projection=Geographic
>  ColumnAmountO3:GridOrigin=Center
>  ColumnAmountO3:GridSpacing=(0.25,0.25)
>  ColumnAmountO3:GridSpacingUnit=deg
>  ColumnAmountO3:GridSpan=(-180,180,-90,90)
>  ColumnAmountO3:GridSpanUnit=deg
>  ColumnAmountO3:NumberOfLongitudesInGrid=1440
>  ColumnAmountO3:NumberOfLatitudesInGrid=720
>  HDFEOS INFORMATION:HDFEOSVersion=HDFEOS_5.1.9
>
> This makes it pretty clear the image is a fullworld image
> in geographic coordinates.  If this is a common metadata
> format the driver could be updated to recognise it automatically.
> Unfortunately there seem to be a variety of ways of storing
> georeferencing for different HDF5 products.
>
>
>  2) Since some part of the tiff is blank, it is not scanned by the sensor.
>> so I want fill them manually,so I wonder the gdal library can do it ? Also I
>> want to add some new tags to the tif and I want to use java to implement it/
>>
>
> Are you wanting to interpolate these nodata areas from nearby valid
> data?  There are many ways of doing this, but gdal_nodatafill.py may
> prove helpful to you as a utility.  If you want to do it from Java
> you will need to call the FillNodata() function which should be
> available in the GDAL Java bindings.  More info on this function at:
>
Thanks, this is just what I want. However I have visite the website of gdal
and the docs, I can not find the gdal java library.
Is it not opened?

>
>  http://www.gdal.org/gdal__alg_8h.html#0a079afef61968f224b159d48423d1c6
>  http://www.gdal.org/gdal_fillnodata.html
>
>
> Best regards,
> --
>
> ---------------------------------------+--------------------------------------
> I set the clouds in motion - turn up   | Frank Warmerdam,
> warmerdam at pobox.com
> light and sound - activate the windows | http://pobox.com/~warmerdam<http://pobox.com/%7Ewarmerdam>
> and watch the world go round - Rush    | Geospatial Programmer for Rent
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/gdal-dev/attachments/20100112/b65b7203/attachment-0001.html


More information about the gdal-dev mailing list