scanned my first book...now what?

noork85 · 05-09-2013, 12:05 PM

hello everyone

ok so yesterday i scanned my first book with the brother ads2000. the whole process was a success that took me roughly two hours. most of that time was spent cutting the book and seperating the pages.

now, my concerns are:

The pdf, it is searchable because i ran it through ocr, still looks like scanned pages. i want to make it look like an ebook with a white background. what do i do?

secondly, i would like to reduce the file size without comprimising the quality of the scanned text.

and lastly, im a complete noob to this process so dont assume anything. please explain in detail what it is that im supposed to do.

thank you!!!

Toxaris · 05-09-2013, 01:41 PM

Search around, this question has been asked many, many times. In short, take the text from the OCR in the wordprocessor you want and fix all the OCR errors. Convert that to HTML and use that as base for your ebook.
Read up on your HTML/CSS skills.

noork85 · 05-09-2013, 06:12 PM

what kind of ocr errors am i looking for?

PeterT · 05-09-2013, 06:50 PM

Quote:

Originally Posted by noork85

what kind of ocr errors am i looking for?

Any / all.
Punctuation
Words not recognized correctly
Double letters such as ll
Formatting issues

mrmikel · 05-13-2013, 03:35 PM

If the book had columns, or pictures inset into the page you want to particularly be watching for sentences which start in one column and end with the text from another. You will have to locate the missing chucks, probably later in the text and reunite them with the text by referring to your book or pdf.

Letter substitutions happen all the time. R becomes E, etc.

Footnotes which may be a good idea in a printed book are much less so in an epub...endnotes at the end of each chapter are better because they are quicker to reach in the same section. This may occasion renumbering as well as links to them and back from them.

Footnotes also, if they continue from one page to another in the same footnote, the second section can end up as nonsense text in the middle of the body of the text.

Even after you think you are done, save the thing and put it into your device and read it again, although probably you will be sick of it by then and you will find yet more errors.

rkomar · 05-13-2013, 06:55 PM

A decent spellchecker helps a lot in finding OCR errors. You still have to go over everything afterwards, but it helps to get rid of most of the errors easily first so that you don't become numb from a flood of them and start missing them as you read.

DaleDe · 05-13-2013, 08:48 PM

Read https://wiki.mobileread.com/wiki/HowTo:_Create_an_eBook in our wiki.

Dale

Stephanos · 05-20-2013, 09:51 PM

As you indicated, your pdf has a text layer from the OCR. One of the more tedious things is getting the text into the word processor without all of the headers, page numbers, etc. that you would rather not have in your ebook. I've found that it is easier to just copy and paste page by page into the word processor. One tip is to hold down the ALT key while you select the text on each page so that you don't get the undesired bits of text.

If you use MS Word, there is a a clipboard feature that will collect up to 24 pieces of text for pasting into Word. So you don't have to keep flipping back and forth between the PDF reader and the word processor.

When you get the text into the WP, it will probably have hard line breaks. This means you have to look at the orginal scan and add an extra paragraph marker at the end of each paragraph. Then, search for double paragraph marks and replace by some placer characters like "~!". Next search for all paragraph markers and replace with spaces. Then search for your placer characters and replace with paragraph markers.

Then you are ready to start the corrections, add back italics, bold, etc. and format to your liking.

Hope this will help you get started. Good luck.

mrmikel · 05-21-2013, 06:28 PM

If it has columns you might do better selecting one column at a time in the pdf reader program. That will keep the columns right. This is time consuming.

Be aware the purpose of creating the text layer is often only for indexing, not to provide the text. For this reason, it is not so trustworthy.

noork85 · 05-24-2013, 01:08 AM

i think im going to have to use the copy and pasting feature.

im not sure i understand the rest...but i guess ill think about that after i do the copy and pasting...

thank you all for your suggestions. keep them coming!

Quote:

Originally Posted by Stephanos

As you indicated, your pdf has a text layer from the OCR. One of the more tedious things is getting the text into the word processor without all of the headers, page numbers, etc. that you would rather not have in your ebook. I've found that it is easier to just copy and paste page by page into the word processor. One tip is to hold down the ALT key while you select the text on each page so that you don't get the undesired bits of text.

If you use MS Word, there is a a clipboard feature that will collect up to 24 pieces of text for pasting into Word. So you don't have to keep flipping back and forth between the PDF reader and the word processor.

When you get the text into the WP, it will probably have hard line breaks. This means you have to look at the orginal scan and add an extra paragraph marker at the end of each paragraph. Then, search for double paragraph marks and replace by some placer characters like "~!". Next search for all paragraph markers and replace with spaces. Then search for your placer characters and replace with paragraph markers.

Then you are ready to start the corrections, add back italics, bold, etc. and format to your liking.

Hope this will help you get started. Good luck.

mrmikel · 05-24-2013, 08:02 AM

What I mean by the rest is that the text layer which is put into some epubs is there only so they can be searchable and is done in a slapdash manner, since anything will do to make the file searchable. So it may be full of mistakes, missing sections, sections out of order. This is why generally if you have any choice, anything is better than a pdf as a source. But often you have no choice.

As has been pointed out before, you will have to proofread line by line to make sure that this has not happened. I have found this in the book I have working on now where an entire paragraph was jumbled...it was all there, but required 5-10 text moves to put it in order.

Notjohn · 06-03-2013, 06:21 PM

The problem may be your OCR software. I bought ABBYY Fineware for about $100 and scanned a novel into Word 2000 in fairly large gulps, say 10-20 pages at a time. To the best of my recollection, Fineware simply ignored the headers, page numbers, etc.

The biggest problem I had was the combination UM in lowercase. I can't remember how Fineware translated it, but it had nothing to do with UM. Since the novel was about ski-bums, I finally had to do a search & replace to fix them all.

noork85 · 06-07-2013, 12:56 AM

Quote:

Originally Posted by Notjohn

The problem may be your OCR software. I bought ABBYY Fineware for about $100 and scanned a novel into Word 2000 in fairly large gulps, say 10-20 pages at a time. To the best of my recollection, Fineware simply ignored the headers, page numbers, etc.

The biggest problem I had was the combination UM in lowercase. I can't remember how Fineware translated it, but it had nothing to do with UM. Since the novel was about ski-bums, I finally had to do a search & replace to fix them all.

the thing is, when i copy and paste into word, everything gets pasted to the side, in short lines. i dont know if im explaining properly. i have to basically press backspace to fill up the line. its very tedious to do so for a whole novel. maybe theres a shortcut, i dont know. im a novice even when it comes to microsoft.

Toxaris · 06-07-2013, 02:05 AM

You can use some intelligent S&R actions. You could also try my add-in, that might help you out.
Also, don't copy paste. Just save as docx from ABBYY (workable copy) and load that into Word.

noork85 · 06-07-2013, 08:21 PM

Quote:

Originally Posted by Toxaris

You can use some intelligent S&R actions. You could also try my add-in, that might help you out.
Also, don't copy paste. Just save as docx from ABBYY (workable copy) and load that into Word.

whats intelligent S&R actions?

and ill get abby today....and try it out...btw, which one should i get? there seem to be multiple abbyy products

05-09-2013, 12:05 PM	#1
noork85 Junior Member Posts: 7 Karma: 10 Join Date: Sep 2012 Device: iPad 4	scanned my first book...now what? hello everyone ok so yesterday i scanned my first book with the brother ads2000. the whole process was a success that took me roughly two hours. most of that time was spent cutting the book and seperating the pages. now, my concerns are: The pdf, it is searchable because i ran it through ocr, still looks like scanned pages. i want to make it look like an ebook with a white background. what do i do? secondly, i would like to reduce the file size without comprimising the quality of the scanned text. and lastly, im a complete noob to this process so dont assume anything. please explain in detail what it is that im supposed to do. thank you!!!

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
sending scanned book to Kindle for conversion?	Georgia Swan	Workshop	0	07-31-2011 10:52 AM
scanned book to epub	langmarp	General Discussions	3	06-28-2010 08:44 AM
Scanned in book only works sideways, or upside down	PGA	Workshop	2	03-12-2010 03:01 PM
Scanned book conversion	jabberwock_11	Calibre	2	01-25-2010 03:37 AM
Google Book Settlement Site Is Up; Paying Authors $60 Per Scanned Book	yagiz	News	8	04-26-2009 01:43 AM

05-09-2013, 01:41 PM	#2
Toxaris Wizard Posts: 4,520 Karma: 121692313 Join Date: Oct 2009 Location: Heemskerk, NL Device: PRS-T1, Kobo Touch, Kobo Aura	Search around, this question has been asked many, many times. In short, take the text from the OCR in the wordprocessor you want and fix all the OCR errors. Convert that to HTML and use that as base for your ebook. Read up on your HTML/CSS skills.

05-09-2013, 06:12 PM	#3
noork85 Junior Member Posts: 7 Karma: 10 Join Date: Sep 2012 Device: iPad 4	what kind of ocr errors am i looking for?

05-13-2013, 03:35 PM	#5
mrmikel Color me gone Posts: 2,089 Karma: 1445295 Join Date: Apr 2008 Location: Central Oregon Coast Device: PRS-300	If the book had columns, or pictures inset into the page you want to particularly be watching for sentences which start in one column and end with the text from another. You will have to locate the missing chucks, probably later in the text and reunite them with the text by referring to your book or pdf. Letter substitutions happen all the time. R becomes E, etc. Footnotes which may be a good idea in a printed book are much less so in an epub...endnotes at the end of each chapter are better because they are quicker to reach in the same section. This may occasion renumbering as well as links to them and back from them. Footnotes also, if they continue from one page to another in the same footnote, the second section can end up as nonsense text in the middle of the body of the text. Even after you think you are done, save the thing and put it into your device and read it again, although probably you will be sick of it by then and you will find yet more errors.

05-13-2013, 06:55 PM	#6
rkomar Wizard Posts: 2,982 Karma: 18343081 Join Date: Oct 2010 Location: Sudbury, ON, Canada Device: PRS-505, PB 902, PRS-T1, PB 623, PB 840, PB 633	A decent spellchecker helps a lot in finding OCR errors. You still have to go over everything afterwards, but it helps to get rid of most of the errors easily first so that you don't become numb from a flood of them and start missing them as you read.

05-13-2013, 08:48 PM	#7
DaleDe Grand Sorcerer Posts: 11,470 Karma: 13095790 Join Date: Aug 2007 Location: Grass Valley, CA Device: EB 1150, EZ Reader, Literati, iPad 2 & Air 2, iPhone 7	Read https://wiki.mobileread.com/wiki/HowTo:_Create_an_eBook in our wiki. Dale

05-20-2013, 09:51 PM	#8
Stephanos Connoisseur Posts: 62 Karma: 133 Join Date: Oct 2007 Location: Minnesota, USA Device: Kobo Aura Edition 2	As you indicated, your pdf has a text layer from the OCR. One of the more tedious things is getting the text into the word processor without all of the headers, page numbers, etc. that you would rather not have in your ebook. I've found that it is easier to just copy and paste page by page into the word processor. One tip is to hold down the ALT key while you select the text on each page so that you don't get the undesired bits of text. If you use MS Word, there is a a clipboard feature that will collect up to 24 pieces of text for pasting into Word. So you don't have to keep flipping back and forth between the PDF reader and the word processor. When you get the text into the WP, it will probably have hard line breaks. This means you have to look at the orginal scan and add an extra paragraph marker at the end of each paragraph. Then, search for double paragraph marks and replace by some placer characters like "~!". Next search for all paragraph markers and replace with spaces. Then search for your placer characters and replace with paragraph markers. Then you are ready to start the corrections, add back italics, bold, etc. and format to your liking. Hope this will help you get started. Good luck.

05-21-2013, 06:28 PM	#9
mrmikel Color me gone Posts: 2,089 Karma: 1445295 Join Date: Apr 2008 Location: Central Oregon Coast Device: PRS-300	If it has columns you might do better selecting one column at a time in the pdf reader program. That will keep the columns right. This is time consuming. Be aware the purpose of creating the text layer is often only for indexing, not to provide the text. For this reason, it is not so trustworthy.

05-24-2013, 08:02 AM	#11
mrmikel Color me gone Posts: 2,089 Karma: 1445295 Join Date: Apr 2008 Location: Central Oregon Coast Device: PRS-300	What I mean by the rest is that the text layer which is put into some epubs is there only so they can be searchable and is done in a slapdash manner, since anything will do to make the file searchable. So it may be full of mistakes, missing sections, sections out of order. This is why generally if you have any choice, anything is better than a pdf as a source. But often you have no choice. As has been pointed out before, you will have to proofread line by line to make sure that this has not happened. I have found this in the book I have working on now where an entire paragraph was jumbled...it was all there, but required 5-10 text moves to put it in order.

06-03-2013, 06:21 PM	#12
Notjohn mostly an observer Posts: 1,515 Karma: 987654 Join Date: Dec 2012 Device: Kindle	The problem may be your OCR software. I bought ABBYY Fineware for about $100 and scanned a novel into Word 2000 in fairly large gulps, say 10-20 pages at a time. To the best of my recollection, Fineware simply ignored the headers, page numbers, etc. The biggest problem I had was the combination UM in lowercase. I can't remember how Fineware translated it, but it had nothing to do with UM. Since the novel was about ski-bums, I finally had to do a search & replace to fix them all.

06-07-2013, 02:05 AM	#14
Toxaris Wizard Posts: 4,520 Karma: 121692313 Join Date: Oct 2009 Location: Heemskerk, NL Device: PRS-T1, Kobo Touch, Kobo Aura	You can use some intelligent S&R actions. You could also try my add-in, that might help you out. Also, don't copy paste. Just save as docx from ABBYY (workable copy) and load that into Word.

Advert

Advert