Accurate optical character recognition at the touch of a button
Itís been a while since I last looked at ABBYY Fine Reader, which now has a much friendlier interface and seems to works far better than I recall.

If you haven't encountered Fine Reader before it's an OCR program that converts scanned documents into editable text. At the same time it can convert them to a range of different formats, including MS Word, Excel, Adobe PDF, HTML and EPUB. It will also handle image files, which may not need the OCR stage, yet it's very efficient at compressing JPGs into PDFs without any perceptible degradation. This could be important if you need to send images by email, because some ISPs won't accept JPGs and may have an upper limit on attachments.

As I mentioned earlier the program's interface is easy to use and has a very Windows 8 kind of styling, with large coloured buttons displaying the available options within a task menu. You can use basic touch gestures to scroll and zoom on desktops or laptops with touch-screens. Obviously to make best use of the product you'll also need a scanner. Having said, that many folk are turning to the latest all-in-one printers, which have one on board. However it's quite capable of working with digital cameras or smartphone files, so it's not an essential element.

Another thing that surprised me was the speed with which the program managed to complete its character recognition taking just a few seconds for a full A4 scan, with very few (if any) mistakes. The layout sometimes needed some attention, particularly when working with forms, but on the whole it produces very impressive results.
For example, I scanned some pages from a glossy magazine, something previous versions had problems with. It's far better with the latest version, converting pages to text and images retaining the layout pretty well, even to the point sizes of the textual elements. There were a few errors, but these tend to be formatting rather than recognition errors. In fact, most of what appeared to be recognition errors were actually words not in the program's dictionary. Users of Microsoft Word will get fewer of these errors, because the program can also incorporate data from Word's own custom dictionary. The main problems I had were with text over graphics and text that was skewed or rotated on the page.

However, if you convert a scan to PDF, the layout is much more like the original which is perfectly understandable, because trying to layout different text sizes adjacent to one another, combined with different spacingís and text over graphics in a Word document is a nightmare.

Other additions in the release include better recognition of foreign languages (190 of them), with full dictionary support for 48 of these. You also now get instant access to scanned documents, while the text recognition goes on in the background.

If you're working with old documents, any post processing can be done within Fine Reader 12's built-in image enhancement editor, so things like photo correction, skew, crop, invert and change resolution can be done without saving the file and loading a dedicated image editor. This same feature can also be used automatically to pre-process scans before they go through the recognition engine. If the documents contain confidential data, you can use the new Redaction mode to hide it. You simply swipe across the text you want to hide using it like a marker, and the text you cover will turn black, yet the output document displays this as a neat row of dots.