mutantstrain wrote:Thanks Tim for your help.
With Omnipage 17, as far as the jpeg conversion goes -- how is that? If I already have Scantailor and Adobe Acrobat (the OCR with Acrobat seems decent enough) , is there really any overwhelming reason to get Omnipage? Seems I can get a pretty good deal on it, but I am most interested in a software product that corrects the usual problems with jpegs: Skewing, cropping etc ... and doing it quickly.
It sort of does that. The reason you still want to feed OmniPage cropped images is that it will sometimes judge parts of the border noise areas as text and waste your time in the editing and correction stages. Scantailor handles that part perfectly. Otherwise it does auto-rotate text for you from upside down to right side up. I'm not sure what process it uses for that, but basically it's automatic and I've never seen it fail. I haven't tested heavily skewed pages, but slight skewed seems fine. Slight page curl is also fine, but you can't just open a book up flat on a desk and try that, because it can't handle that much curl. It's other advantage is the batch aspect of the processing. You can open a set of images and OCR them in not many clicks or keystrokes (I think 2, one to OCR and one to save the results if you don't want to use it's editing features.). It also has flatbed scanner hotkeys if you do that as well, but I assume Adobe does as well.
I can't say if Omnipage would offer any advantage over Adobe for you, since I've never used that for more than a page or two. I don't know if Adobe's OCR is trainable, but that is something you get with an OCR package like OmniPage or ABBY Finereader. You can train a new font or character, etc to reduce the error rate on a book. If you already have Adobe's OCR, another may not be worth it for you. OmniPage and Finereader seem to work about the same, but I'm only familiar with the accuracy of OmniPage.