ABBYY FineReader Express 8.4: recognize text from any source on the fly (distribution code completed)

Digital content, and electronic versions of documents surround us on all sides. Paper in our lives is almost no space left. Newspapers and magazines have moved into the online format of the book we read on e-ink reader or tablet, ordinary letters have replaced email and sms. Yet, sometimes we have to tinker with the first papers to get them an electronic copy. Here we come to the aid of special programs that use OCR technology to OCR text (Optical Character Recognition). The most famous of these is undoubtedly ABBYYFineReader. You can use it to convert paper documents into editable formats, and save the PDF to searchable text. And today we have a great opportunity to learn more about it.

* * *

For Macs, ABBYY offers only FineReader Express, it nevertheless has the necessary functionality. The key features of ABBYY FineReader Express is a recognition accuracy and layout retention, support for many languages ​​(171 language to the three languages ​​in one document), the transformation and the creation of PDF (PDF conversion to editable formats), editor for manual marking areas (text, table, picture) and a simple, user-friendly interface programs.

first look

FineReader Express operating window is quite minimalistic, there exist only the most necessary items. The side panel contains sketches added pages and on isntrumentov panel buttons with drop-down lists to select the language and output file. Still there is the conversion and the zoom button. Otherwise, the interface corresponds to a fast express version, which bet on the automatic execution of operations with a minimum of configuration and user participation.

Pass the tests

After launching FineReader meets us a compact window with a choice of scenario. Here we are asked to select the capture source: scanner, fax, or read from a file. It is also advisable to specify a document language (or languages, if more than one) - it will help to improve the recognition accuracy of the original document. Well, actually the output file format, everything is simple - choose based on the type of the paper document.

Scanner at hand I did not have, but it's even better - using as the source of the photo Made with the help of the iPhone, I complicated the task to recognize text. As an example of the text, I took one of the books of his wife, as well as an example of the table - some old working film consignment of iPhone. Well, let's get started.

Each page with text

For lack of a scanner I just did a book turn photos - photo normal room light, no tripod, and other tweaks. Here is the original:

Let's see what it can do with FineReader. We specify that we want to pull out the photo text, define the language like Russian, and start the process.

To its credit, the application must be said that the entire text was defined, including accidentally got to bend the adjacent page. A piece of the table, which I specifically left in the frame, defined as the expected picture. But it's not scary, because we can manually change the domain, specify its type (if the program is not set correctly) and remove the field, the recognition of which is required. All manipulations took me less than a minute, but in the end I got here is a quite acceptable result:

After a short proofreading and edits the document is ready. I think this is a good result for such a quick, almost automatic recognition process.

recognize table

As an experimental table serves unpretentious bill, which was also filmed on the iPhone. There is already in use Ukrainian (along with check language support), which is also good for our experience. Choosing a new script (⌘N) Indicate the source - read from a file, the language - Ukrainian, and file output - table.

The program thinks for a few seconds, and here we have the result:

With the table program is not handled so well, but it is more or less acceptable, in principle, the text of the definition, unless the reason to finish the cells that were not in the original document. There will have to tinker a little longer to get the final form of the map document, but it is easier than typing a sign with the hand from scratch.

Save to PDF

When saving to PDF, the program unfortunately does not improve the original image (contrast, brightness) and it is placed in the PDF-document as is. But the less, the search text is present, and that's good.

Total

Like any tool, FineReader has its pros and cons. The strong points, in addition to the stated characteristics of the manufacturer, is that the OCR tables and works quite well, and convert to PDF, as promised, supports search text. The downside is the lack of options and very meager means for manually controlling the process. But this is partly justified, the fact that it is an express version and it works automatically.

Codes for FineReader Express program won Gregory Ushar and Nikolai Blinov. Congratulations! Check your private messages, codes sent.