Jonathan,
The dem_mosaic utility in the NASA Ames Stereo Pipeline suite of tools does 
exactly what you want.  It was developed to efficiently generate large seamless 
mosaics from 1000s of raster inputs.  The user can select from several options 
for output mosaic values (weighted average, median, first, last, stddev, count, 
etc.).  I recommend that you start with the default weighted average option.  
Documentation, binaries (Linux and OSX), and source code are available here: 
https://ti.arc.nasa.gov/tech/asr/intelligent-robotics/ngt/stereo/.  
-David




---
David Shean
Polar Science Center, Applied Physics Lab
Dept. of Earth and Space Sciences
University of Washington
Box 355640
1013 NE 40th St
Seattle, WA 98105
PSC website <http://psc.apl.uw.edu/people/post-docs/david-shean/>

Office: (206) 221-8727
Cell: (301) 801-4558
Fax: (206) 543-6785

> On Feb 6, 2017, at 2:33 PM, Jonathan Greenberg <[email protected]> wrote:
> 
> GDALers:
> 
> I love the gdalbuildvrt -> gdal_translate trick for large mosaicks, but I 
> have a somewhat more complicated problem I was hoping I could get some help 
> on.  Given a large set of overlapping images, I want to mosaic using the 
> following "rules": 1) all overlapping pixels should be averaged together, and 
> 2) any "nodata" pixels in a given input image should be ignored in that 
> calculation.
> 
> Is this doable in GDAL?  Or am I stuck with just the "top" image in the vrt 
> being the value the output mosaic uses?
> 
> --j
> -- 
> -- 
> Jonathan A. Greenberg, PhD
> Randall Endowed Professor and Associate Professor of Remote Sensing
> Global Environmental Analysis and Remote Sensing (GEARS) Laboratory
> Natural Resources & Environmental Science
> University of Nevada, Reno
> 1664 N Virginia St MS/0186
> Reno, NV 89557
> Phone: 415-763-5476
> http://www.unr.edu/nres <http://www.unr.edu/nres>
> AIM: jgrn307, MSN: [email protected] <mailto:[email protected]>, Gchat: 
> jgrn307, Skype: jgrn3007
> _______________________________________________
> gdal-dev mailing list
> [email protected]
> https://lists.osgeo.org/mailman/listinfo/gdal-dev

_______________________________________________
gdal-dev mailing list
[email protected]
https://lists.osgeo.org/mailman/listinfo/gdal-dev

Reply via email to