<br><font size=2 face="sans-serif">Hi,</font>
<br>
<br><font size=2 face="sans-serif">I've been looking at using gdal_translate
to convert an HDF4 image to geotiff. During my investigation, I happened
to notice that the gdal geotiff driver might not be handling the "pixel
is point" verses "pixel is area" distinction properly.</font>
<br>
<br><font size=2 face="sans-serif">If I understand things properly, the
gdal default for pixel coordinates is the same as geotiff "pixel is
area". The upper-left image corner specifies the upper-left
corner of the upper-left pixel, and the bottom-right image corner specifies
the lower-right corner of the lower-right pixel of the image.</font>
<br>
<br><font size=2 face="sans-serif">According to http://www.remotesensing.org/geotiff/spec/geotiff2.5.html#2.5.2,
the "pixel is point" option changes how the coordinates in the
geotiff file are interpretted so that the image corner points are relative
to the center of the pixels instead of the edges.</font>
<br>
<br><font size=2 face="sans-serif">I don't see anything in the geotiff
driver that is handling the "pixel is point" case to properly
fix up the affine transformation coefficients. Am I missing something?</font>
<br>
<br><font size=2 face="sans-serif">Also, it would be nice if a "pixel
is point" option could be passed to the geotiff driver when generating
a geotiff output. (Assuming my interpretation of pixel is point/area is
correct). That way the corners passed to the process could be relative
to the center of the pixel if that is what is currently available.</font>
<br>
<br><font size=2 face="sans-serif">Thanks</font>
<br><font size=2 face="sans-serif">Tim</font>
<br>
<br><font size=2 face="sans-serif">--------------------------------------------------------------------------<br>
Tim Beckmann<br>
Software Project Lead<br>
Science Applications International Corporation (SAIC)<br>
Contractor to the USGS/EROS<br>
Sioux Falls, SD 57198<br>
605-594-2521 Phone<br>
605-594-6940 Fax<br>
tbeckman@usgs.gov<br>
</font>