How are template images used in ADR?
Posted: Sat May 19, 2012 4:48 pm
I understand that images are added to a template to assist with detection, but I have questions to determine WHICH images to addd to a template. I ASSUME it would be best to use slices / snips of an example document... which means possibly constructing a simple UI to allow a user to disect an existing regions and then extracting those regions as images for use in the template. However, these are just assumptions. Could you help clarify these ideas? More specifically...
How does one decide which images to add to a template for ADR?
How are those images used when performing the comparison?
Should the images be whole scans of sample documents or key anchor points?
Should the images contain text, graphics (logos), or both?
Thanks, in advance,
Fred
How does one decide which images to add to a template for ADR?
How are those images used when performing the comparison?
Should the images be whole scans of sample documents or key anchor points?
Should the images contain text, graphics (logos), or both?
Thanks, in advance,
Fred