[bksvol-discuss] Re: images scanning and recognizing

  • From: "Gary Petraccaro" <garyp130@xxxxxxxxxxx>
  • To: <bksvol-discuss@xxxxxxxxxxxxx>
  • Date: Wed, 5 Jan 2005 01:01:06 -0500

Thanks for the explanation of the prefix with images.  Had forgotten all about 
that.

  ----- Original Message ----- 
  From: Pratik Patel 
  To: bksvol-discuss@xxxxxxxxxxxxx 
  Sent: Tuesday, January 04, 2005 9:11 AM
  Subject: [bksvol-discuss] Re: images scanning and recognizing 


  E,

  The best way to accomplish what you want--especially in terms of multiple 
book images--is by:

  1.  get your settings perfected (either by using the  "optimize feature" or 
by using the Gray Scale 400 DPI method.)  More on this later.

  2.  Go to the "settings menu"

  3.  Go to "scanning" by using your arrow keys.

  4.  you will land on "mode."   Your current setting will be "scan and 
recognize."  Change this setting to "scan image." 

  5.  Tab over to "image prefix" option.  Type an appropriate prefix.  E.G. If 
you were scanning a book such as Complete works of Vivekananda, you may want to 
ocnsider tping CompleteVivek" as your prefix.  This will automatically add the 
prefix you choose to the image numbers that K1K scans.

  6.  Tab to OK and press enter.

  7.  Go back to settings" menu.

  8.  Choose "save settings" (v).

  9.  give your file an appropriate name such as "CompleteVivekananda."  This 
will save a settings file with your prefix  in it.  Press enter when done.

  10. Scan away.  When done, you can leave it or follow the steps to scan 
another book.  When you create your optimal settings, make sure you go back to 
the settings menu and change the "mode" to "scan and recognize."


  11. when you feel like recognizing the images you scanned, go to the settings 
menu ahd choose "load settings."

  12. Arrow up and down to load the settings file you saved earlier.

  13.  Go back to "settings" and "scanning."  Choose the mode to "recognize 
image files," and press enter.

  14.  Either press F9 or go to the "scan" menu and choose "scan a page."  This 
will begin recognizing images that you scanned with the particular prefix you 
stored in your settings file.  Once all images have been recognized, you can 
save the file as you would any other file.

  15.  Follow the procedure for all the other images you scanned with different 
prefixes.

  I hope this saves you some looking up.

  Now as to your second questions regarding Gray Scale and 400 DPI--The 
procedure that I follow is that, when I begin scanning a book, I generally 
optimize my settings with the option in K1K.  I save the settings that K1K came 
up with in a settings file called "current."  This allows me to refer back to 
it when necessary.  Then, I open the book up at a random page and scan about 10 
consecutive pages to gage how the scan came out  with the settings.  Once all 
the pages are done, I use the rank spelling feature to rank the OCR mistakes.  
Once I have an idea of the amount of mistakes EtC, I change the scanning 
setitngs to the Gray scale/400 DPI and rescan the same 10 pages that I scanned 
previously and use the rank spelling feature to gage the outcome.  i choose the 
better setting to scan my book.  I find that Gray Scale with 400 DPI happens to 
be  the setting that gets the best results most of the times--especially when 
it comes to paperbacks.  There are, of course, exceptions.  Properly settings 
up each scan saves a lot of work in the end when the final product comes out.

  Hope this helps.  Let me know if you have any questions.

  Pratik

  Pratik Patel
  Interim Director
  Office of Special Services
  Queens College
  Director
  CUNY Assistive Technology Services
  The City University of New York
       ppatel@xxxxxx

  -----Original Message-----
  From: bksvol-discuss-bounce@xxxxxxxxxxxxx 
[mailto:bksvol-discuss-bounce@xxxxxxxxxxxxx] On Behalf Of E.
  Sent: Thursday, December 30, 2004 8:54 AM
  To: bksvol-discuss@xxxxxxxxxxxxx
  Subject: [bksvol-discuss] images scanning and recognizing

  Is there a way to scan but not recognize several books at one sitting and
  have each set of images saved in a separate file for later recognizing.  If
  so, please explain how and how the recognized file would be saved to disk
  once the images were recognized for each book.  I am sure this is in the
  manual somewhere but asking is quicker.

  Yes I have read the manual and plan to re-read it.

  A second question.  I notice some of you seem to use gray-scale and 400 dpi
  even when the optimizing feature does not suggest it.  Please comment and
  comment on how to find out which gives the best statistics.  How do I clear
  my recognision statistics between books?

  E.



Other related posts: