I have a couple hundred historic air photos for a county. When I add all the pictures into arcview, the result looks alright, the collars hide parts of neighboring pictures. If needed I could turn a particular picture on and off so that I can view what is behind it. However that takes time, and also doesn't work for online mapping applications. I also would like to eventually export the entire set of air photos into a single image. Normally I would set true black as transparent for Landsat, or other remote sensing images. However, each of the photos has a black collar around them which appears black, but is not "true black" and also it is not necessarily the darkest part of the picture. Therefore I have to figure out something else.
I think it might be possible to create a mosaic and use the historic air photo that has the closest center point to populate the value of the new raster. For example in the image below the black collar of BJC-4-20 covers part of a field south of town. The field does show up on image BJC-4-46. Since that field is closer to the center for BJC-4-46, I would like to use that historical photo as a source for the new raster, and effectively remove the collars from the new air photo.
Does anyone have any other ideas on how I could remove the collars off a bunch of air photos, and create a single new air photo image?
Depending on the amount of overlap that you have, you can create a mosaic dataset and then use the Build Footprints tool and then define a shrink distance enough to cover the fiducials that you want to remove. It may take a little of testing to determine the correct shrink distance, but since the Build Footprints tool offers the ability to reset the original footprints it should not be too bad.