My customer insists on the OCR of a single image file containing 1-5 48 page booklets (each containing 760 checkmark groups), plus 1-2 8 page booklets (each containing 125 checkmark groups), plus x pages of supporting documentation that does not contain any data. Initial testing is resulting in an unacceptable number of page recognition errors. Is attempting to apply two document definitions to this image feasible, or would we be better suited scanning each booklet into a separate image file?
