[gdal-dev] Building TIFF overviews from bitmap in memory

[yodan]

Hi, 

I have a large bitmap in memory, and I need to build overviews for this
image and save them in TIFF format in an .ovr file. I could do this by
saving the bitmap to disk (as a TIFF file), and then calling BuildOverviews
with the path to the TIFF file, but this would use too much unnecessary IO
which I cannot afford.

Now, I figured out how to use VSIFileFromMemBuffer to open my in-memory
bitmap in GDAL without writing it to disk. However, according to my
understanding of BuildOverviews (and correct me if I'm wrong), it will use
the same driver that read the input to produce the output. I understand this
means that when building overviews for a bitmap, the .ovr file will store
the overviews as bitmaps.

Is there any way to build overviews in TIFF format for a bitmap input,
without having to convert the large image to TIFF format?
Alternatively, can I perform the downsampling myself and only let GDAL
manage the format of the .ovr file (by supplying the images for each pyramid
level)?

Thank you for your help!
Yodan

(C++, Windows, MSVC 2012, GDAL 1.9.1)



--
View this message in context: http://osgeo-org.1560.x6.nabble.com/Building-TIFF-overviews-from-bitmap-in-memory-tp5202974.html
Sent from the GDAL - Dev mailing list archive at Nabble.com.