2009/4/23 John Jason Jordan <[email protected]> > On Wed, 22 Apr 2009 20:52:29 -0600 > Walter Hildebrandt <[email protected]> dijo: > > > Does anyone know if there is a hand held copying device that can copy > text > > and paste that text into a Writer text file and be able to edit that text > in > > the Writer text file? > > There used to exist such items a long time ago. I don't know of any > still being manufactured. The problem was not the scanning technology > but, rather, the unreliability of the human operators. It turns out > that it is very hard to hold the device steady enough. > > Note that there are two functions involved to achieve the results you > desire. The first is scanning, which converts the reflected image from > the page of text to a bitmap image. The second is the optical character > recognition (OCR) function. OCR is a process where software looks at > the bitmap image and recognizes letterforms. > > The scanning function today is extremely accurate with modern scanners. > The OCR function is where things fall apart a bit. Think about those > web sites where you have to enter the distorted characters from a > bitmap image into a box to be sure you are a human and not a spambot. > That's the problem the designers of OCR software face, and it's not > trivial. OCR software typically trips up on distinguishing a zero from > an o, an s from a 5, and so on. Even the best OCR software used on the > cleanest text misses quite a bit. Plan on having to clean it up. >
In addition to the above there's a third issue. By far the majority of scanners and OCR "engines" on the market produce documents that are designed for MS Office. Many will even invoke MS Word or Outlook automatically once the OCR process is complete. Of course that's a problem if you don't have either of those installed - which I don't. My scanner wanted to produce a ".doc" file and then automatically invoke Word. I can't separate those two functions. In other words I can't make my scanner produce a ".doc" file and then *not* invoke Word. I certainly can't get it to invoke Writer. The best I can do with my scanner (HP 3770) is to make it scan to an RTF file and then not invoke anything automatically. This is OK but RTF is a lot feature-poorer than ".doc" so my results look a lot less like the original than they could and therefore I have a lot more work to do. Of course most OCR software will let you process an already scanned "image" file so you can separate the scanning from the OCR and, for example, acquire a better OCR package than the one that came "in the box" with your scanner. My OCR package (Readiris Pro 9) will actually let me scan to OOo 1.0 or 1.1 or to StarOffice 6. So perhaps if I upgraded (they are on version Pro 12 now) I might be able to get software that will save directly to OpenOffice 2.x or even 3.x or at least to a current version of ODF. -- Harold Fuchs London, England Please reply *only* to [email protected]
