I have 96 aerial photos scanned from film that I need to geo-reference and mosaic. Each photo has a border and image numbers that need to be clipped for the mosaic. They were scanned at a high resolution in tiff with a file size of about 900 mb each. Ground resolution is 1 ft. The photos are along a river corridor and so the north orientation is different for each photo. I'm using ArcGIS Arcinfo service pack 4.
I'm wondering about compression or resampling, clipping the border, and the most efficient work flow to get this done. What would you do?
File storage is not really a problem but the images are slow to draw. Should I build pyramids and use original files, compress the files, before or after georeferencing?
Should I clip the borders before or after georeferencing? If I use a mosaic dataset, can footprints be used to address the borders, but do I have to set up a footprint for each photo or can they be batched? Will the different rotations of each photo make using footprints more challenging? Should I try clipping photos in photoshop before georeferencing? If so can they be batched?
Any advice will be be greatly appreciated. I'd rather do this right from the beginning rather than start over after going down the wrong road.
Thanks,
Jeff