Jump to content
Science Forums

OCR experience


Recommended Posts

Recently I tried using the OCR software that came with a Cannon scanner to transfer some lists of numbers to a text format. Miserable! No matter the format (color, grayscale, B&W) or resolution (150, 300, 600), the software identified correctly less than 3% of the charachters. Anyone having any better experience with OCR.

 

PS I had to settle with jpg files on my page:

http://home.comcast.net/~turtlediable/wsb/index.html

:)

Link to comment
Share on other sites

I've had 95% matches on older version of OCR, but the picture was really clear and neat, but i think it would be easier to type up the papers, yes it will take a little time, not saying it wont, but it will be time well spent if you know what i mean... Actually before you do that, i'll ty to go another way tonite, I will get the paper, open it in photoshop, select all letters on the page, color them black, to be sure they are, and then try running it through OCR and see what happens...

Link to comment
Share on other sites

hey if you wanna do any type of complex equation typing, you want to use latex, a software that is 100000 times better than its rivals, also its free and open source, and there's quite a large amount of documentation on it, its really awesome, i use it for all papers that i do, but their math library is the best!

Link to comment
Share on other sites

If you want results here's what you do:

Get Abby FineReader OCR Professional (preferably new version, and you can trial it, since you dont need to do that many documents. P.S. its 39 meg download). Install and start Abby, Open your jpeg, now, dont click recognize on the bar on top, what you want to do is go into process and first do analyze layout after that do process read, and you should have similar results to the ones posted.

 

P.S. if you dont, what i did is i went into photoshop and changed contrast and brightness to 100%, and thats the image file thats included...

 

P.P.S. Tormod, it would be nice to include rtf files as uploadable too, RTF is Rich Text Format, and should not present problems as it is just text, also consider openoffice format...

 

P.P.P.S. If you want to see the documents, send me an email to [email protected], and i'll send both the image i used and the rtf, since both go beyond the restrictions of the attatchment size...

Link to comment
Share on other sites

  • 6 months later...

___Wow; it's quiet in here. Frankly I'm surprised it's not closed. Echo..echo..echo.ehco :)

Anyway, did I ever mail you Alexander? Prolly not; I think I just typed it up. Which brings us back round to the topic of OCR. Here I am again with half a dozen pages of math, this time handwritten. No way huh!? We can blow a rocket up in space but can't read our own writing. This is not even arcane notation, just printed numerals, letters & keyboard punctuation like braces; no exponentiation, no summation.

___I will put up a scan of the first page in Katabatak thread (this is where I intend to use the list anyway) & make it as big & clear as allowed. Somehow I don't thing clarity of image is a problem this time so much as handwriting.

___Sorry if I made too much noise. ;)

Link to comment
Share on other sites

I think that this time you'd be better off typing it up. Plus Latex notation is really great, i think that for the time it takes you to spend on learning latex, you'd get a lot more in return. There are interfaces that allow you to use latex on mac, I'm not sure about Linux or if there is even a port to Windows as of yet...

Link to comment
Share on other sites

Anytime, Turtle :circle:

 

as to the closing of threads, i dont tend to close them anyways, people sometimes post on year old posts and still they bring something new to the topic which makes it interesting to discuss, yet at the same time if it was posted as a new post there would have to be much to be explained...

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...