LEADTOOLS OCR – Text Recognition free on the Mac App Store
A time to perform optical character recognition of a printed page is resorted to very complex and often expensive software, today there is a free app on the Mac App Store called LEADTOOLS OCR.
The challenge ourselves aware that we cannot expect the same results as reference applications for the category, but being free, we are ready to accept some compromises in terms of functionality.
Just open the interface is very streamlined and easy to understand, with a thumbnail of the document to the left, the largest open in the center and a toolbar with functions. Give it a try with the test document already preloaded (in English) and gives us good recognition results.
We choose File / Open and open a PDF document using a scanner and just got the app preview, captured at 150 dpi from the page of a book. The text looks a bit soft, have no boundaries especially engraved, but let’s try the same to complete the operation. First, we use the button at the top right ” Find Zones “, which automatically identifies and beads of red areas of the document that contains the text.
Once you have identified the areas of text we click on the ” Recognize Text ” and we will ask the name of a text file in which to save the result. With our PDF test we found numerous errors of recognition, one or two for each row. Obviously, this is a poor result that requires a thorough manual correction, although it may be convenient to be with the bulk of the text ready.
With the test document in English LEADTOOLS OCR proved to live up to expectations, while confronted with a document scanner and just acquired in the other language has shown several limitations. Apparently, the main problem is the quality of the image that it provides, because if this is soft and lacking in contrast it undergoes a large amount of errors.
Improving the image source, I have complied a partial reduction of errors, but the result always requires a significant corrective surgery. All in all, it may be sufficient to use sporadic views of the gratuity, but not exceed, the sufficiency of our evaluation.
