Is there a way to clip the edge of an imported map to remove a border which was made during scanning

Question

Using arcmap 10.4.1. I am recreating an old map that has been scanned in as individual tiff tiles.  I have imported them, georeferenced them etc and am now trying to put them together to form one map.  Some have borders created when they were scanned and I need to get them to butt up against each other. At the moment they are overlapping.  I would also like to make them all greyscale as each tile is a slightly different colour,

JaySul · Answer

Thank you for getting back.  I will try out the mosaic dataset - not used this before as I have only done basic work in GIS.  I did try to take each image into Gimp and clipped and recoloured it but when I put it back into GIS I was not able to open its properties to set the spatial reference -  I got the message  'failed to open dataset'.

The colour of the tiles range from pale cream to pink and everything in between.  I want to find a uniform colour so that it looks like a composite map when I reproduce it. I thought greyscale would be the simplest as otherwise I would have to alter each tile to get them to match up.  I will try to attach a screen print of what a section of it currently looks like.

screenshot Beighton.png

DavidPike · Answer

A simple way without having to alter or recreate the tifs is to create a mosaic dataset, add the rasters to it then create footprints.

The footprints assigned to each raster essentially control how much of a raster is to be displayed.  You can create them radiometrically if your raster border has common pixel values, or you can edit the footprints post facto which is what I'd probably advise (unless you have 1000s to do of course)

https://desktop.arcgis.com/en/arcmap/10.3/manage-data/raster-and-images/creating-a-mosaic-dataset.htm

https://desktop.arcgis.com/en/arcmap/10.3/manage-data/raster-and-images/wkflw-replacing-the-footprints-in-a-mosaic-dataset.htm

why greyscale? are they tru-colour at the moment? the mosaic dataset also has options for dodging and colour balancing which i'd certainly try.