We have one very large contiguous area of raster. To speed up the processing time we split the area into a number of areas of raster and have a separate folder for each area . Obviously, the folders are all contiguous and match up with each other and for most of our data there is no overlap and they match up perfectly.
We ran the 1, 2 and 4m on each folder creating 1m, 2m and 4m AT5s for each folder. It appears after taking a look at where the black areas are that they are all appearing at the borders of where the folders would join. After thinking about this, I believe it is likely occurring because each IMC is creating its own block list for its own folder area and there are blocks outside the raster for that folder of images. So when you take the at5s and join them all up it still sees the black where the folders join. If all of the folders were in the same IMC to run an AT5 level or if all of the photos were in one folder this probably would not occur.
To avoid this can anyone advise with certainty .....is the simplest way to just take all of the folders and put them into a single IMC to run each level? From a time perspective, I was hoping to avoid that because we have lots of folders. If I do this and but still then get black borders at the outside edge of the raster area I assume I can then back fill that area with some additional raster, and run the at5 on just on that fringe area and add it to our other AT5s to avoid the black.
Is this the best way to proceed?
Does anyone know if you have say 100,000 contiguous photos if the IMC would process it faster if they were all in a single folder vs. 20 loaded folders each with 5,000 images or is it the same? We are just trying to figure the most efficient way to use the software to run our data. We were able to get all it run pretty quickly running lots of IMCs on lots of folders but that does not appear to be possible without creating black borders at the edges of each folder. Thanks!