OCR's a hit & miss business, when it works it's bloody marvellous, but when it doesn't it drives you mad. I've tried a few freeware ones like 'TopOCR' & 'SimpleOCR', both good for sharp clear text but fall over completely on even quite good quality photocopies or original documents. Been thinking of getting a scanner upgrade anyway, but has anyone used any particularly good OCR programmes? Wondering also if anyone perhaps has a proprietary one that came bundled with a scanner that's proved good on the more difficult documents? ~A
My Canon scanner is now 5/6 years old, but it was bundled with Pictbridge, which I have found pretty good. What it must be like now, 6 years later, with a latest version, I can only but guess at!
Adam I've been using http://www.abbyy.com/finereader_sprint/ for years now but can't remember which printer it came bundled with I see that they do a free trial....... worth a shot ? I see you say OCR's a hit & miss business, when it works it's bloody marvellous, but when it doesn't it drives you mad. Ain't that the truth Regards Ron
VP, I like Ron have been using the Abbyy Finereader that came with my Canon 4400. I liked it so much, I bought a second one to use at work. It will do slides. The OCR is teachable. I love the 1200 dpi it does with pictures. This unit is a few years old and there are probably better ones out now, but I think a dedicated scanner instead of one of these scanner, printer, etc is the ticket. cheers, phil
It just occured to me that there may be some members on the forum who have yet to succumb to the delights/exasperation of OCR. There are plenty of explanations of OCR (Optical Character Recognition) on the internet but if you want a simple and quick intro read on: If you have an image on your PC (say a jpg.) that shows a document then it will always be just an image and you cannot edit the text. What OCR does is to scan the image and every time it identifies a character it will store that info in it's correct place. On completion the program will reproduce the whole image as, for example, a WORD document which can then be edited exactly as any other document . There is much more to OCR than that, of course, but that's enough for you to get interested.
Some times it works and *7Rw2 JkkR;' d87*6-244,;0= It dON'T! Don't I just know that feeling... Cheers for the suggestions so far chaps, I suppose like most computer things it'll get better and better. Kind of staggers me that we can even do a crude version of it at home these days at all. ~A
Been using Omnipage pro 14 for some time now in conjunction with my Lexmark All In One printer/scanner, and up to now cannot fault it even though there are more up to date versions available. Regards Peter
Abby FineReader works well for me, but I had to buy it. The secret with them all, I suppose, is scanning on high resolution to decrease the chance of error, and then selecting the right scanning language. FR will read scans, files kept, PDF files, etc.
Ron; Cheers for the synopsis. I'd been avoiding this thread simply because I didn't know what it was about. Having looked, yours is indeed the only post which makes it any clearer
Must agree about ABBY Finereader, it's a great OCR programme. I'm using the Pro version which has a number of different languages built it. Funny enough it seems to recognise German text better than English (as long as it is not in a Fraktur typeface, of course). The thing that really drives me crazy with OCR software is when they do a good job of reading a bad original and then make stupid mistakes in an excellent original. I've never figured that one out!? But to be honest, if your original text is not perfect then you're still better off learning to touch type then waste your time with OCR; epsecially if there is a lot of formatting that needs to be kept or changed. For example if I have twelve photocopied pages of an old typewritten document, I can re-type them faster and more accurately than scanning them, doing the OCR and then correcting all the mistakes. Where it works well and is a benefit is when the original documents are professionally printed, first generation copies. Lee
Hi Lee Nice to see you keeping in touch. I'm terribly sorry, but on spotting your contribution I just couldn't resist posting this ! "But to be honest, if your original text is not perfect then you're still better off learning to touch type then waste your time with OCR; epsecially if there is a lot of formatting that needs to be kept or changed" I glad to see your touch typing is no better than mine Best regards Ron
:icon_smile_blackeye You got me there Ron, that kind of destroys my argument! Hi Lee Nice to see you keeping in touch. I'm terribly sorry, but on spotting your contribution I just couldn't resist posting this ! I glad to see your touch typing is no better than mine Best regards Ron
Mine came with cheap printer scanner that a friend gave me. (he already had a better one and it came with the new computer) It works very well indeed reading and deciphering worn out war time documents. What I want is a Voice recognition to text software that works. I have "Dragon" and it dont! Sapper
I tried to use my OCR program yesterday after reading this. Using my Lexmark 2600 series priner/scanner. I scanned a letter from a bank as it was nearest thing to hand. I must say it was rubbish & then I couldn't edit it as I don't know my 25 digit code to get Word working & can't find it anyway. Still if I (or rather the kids for school) need a WP prog I can use Works WP & the scanner works fine usually as you can see form alot of my posts on here.
I have never possessed a programme as outlined on this thread, but the idea sounds really good if it functions. It really sounds from members personal experience that a lot still do not work as promised. Perhaps I will stick with my steam driven methods for the time being at least Regards Tom
you can use abbyy for free (10 pages per day) online. never tried it myself, but could be worth a go... ABBYY FineReader Online
Keep trying. Only by experience do you get the skill to recognise what will copy and what wont. Low res wording simply will not register so you might have to mess about with the contrast on the original. Sometimes I have to scan a jpg, alter the contrast and maybe re-size it. Then print it out before re-scanning it to make a copy that is good enough for OCR!. I have to say 75% of the time it works just fine but even the best have a number of errors and you HAVE to check it with a line by line read. You can not scan a document that has more than one column. If a page has 2 columns then it must be done one column at a time. Page titles and nunbers are also to be avoided.
