Pisa tower syndrome

roger64 · 11-15-2011, 11:11 PM

Hi,

I use Fine Reader (v.9) to ocr some PDF (images) files of old books. The results are usually very very good. But sometimes, say in a 100 pages document, some pages are not vertical but slightly inclined on one side (some few degrees are enough). Yes, like the Pisa tower... Then the ocr result worsens noticeably.

Is it possible to modify some pages of a image PDF document to correct them and make them look vertical?

Which software (preferably free) may I use to do it?

DDHarriman · 11-18-2011, 05:12 AM

Hello

Yes, and one can do that in several ways.

One way can be by exporting the culprit pages as images - per example tif non compressed format.
Than correct the image in an image editor like Photoshop - a free and very lightweight alternative is Pain.NET (http://www.getpaint.net).
Import the edited pages back into finereader, and re-ocr all the images (or if you have ocred all the other pages ocr just the imported ones).

Another way can be not to scan directly with finereder (this I think is what you are doing, right?), but to numbered image files with the scanner own application - probably you will get better success rate this way.
Than, if one or more pages are not straight, use an image editor to correct them.
Import all the images into finereader and ocr them away.

Best regards,

roger64 · 11-18-2011, 06:12 AM

Hi,

I'll try the first solution because I have no scanner. I download my PDF image documents from Gallica (Bibliothèque nationale de France).

But it would be about time I buy one...

Thanks for your tips.

DDHarriman · 11-18-2011, 08:05 AM

Hello

If you can afford it, try to buy a dedicated book scanner.

My advice is the Plustek Opticbook 3600 (http://plustek.com/usa/products/opticbook-series/).

Best regards,

DSpider · 11-21-2011, 09:05 AM

I don't know about FineReader 9 but version 10 has an "Edit Image" button with some very interesting options. You should check it out. But I supposed you could run the images through Scan Tailor, then OCR them. Just make sure they come out grayscale and not in black and white, for more accuracy (because FineReader already has filters in place).

11-15-2011, 11:11 PM	#1
roger64 Wizard Posts: 2,608 Karma: 3000161 Join Date: Jan 2009 Device: Kindle PW3 (wifi)	Pisa tower syndrome Hi, I use Fine Reader (v.9) to ocr some PDF (images) files of old books. The results are usually very very good. But sometimes, say in a 100 pages document, some pages are not vertical but slightly inclined on one side (some few degrees are enough). Yes, like the Pisa tower... Then the ocr result worsens noticeably. Is it possible to modify some pages of a image PDF document to correct them and make them look vertical? Which software (preferably free) may I use to do it?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Free (Kindle) Syndrome by Thomas Hoover	arcadata	Deals and Resources (No Self-Promotion or Affiliate Links)	2	08-31-2011 01:21 PM
Finally: Relief for Restless Leg Syndrome	WT Sharpe	Lounge	8	04-06-2011 11:06 AM
Science Fiction McDonald, Steven E.: The Janus Syndrome , v1 28 Jun 2008	Madam Broshkina	Kindle Books	1	02-13-2011 09:01 PM
Science Fiction McDonald, Steven E.: The Janus Syndrome , v1 18 Jun 2008	Madam Broshkina	BBeB/LRF Books	0	06-18-2008 08:48 PM
Hello from Pisa, Italy	giamba	Introduce Yourself	2	05-14-2007 06:40 PM

11-18-2011, 05:12 AM	#2
DDHarriman Guru Posts: 860 Karma: 4380 Join Date: Feb 2008 Location: Almada, Portugal Device: Cybook Gen3, Sony PRS 505, Kindle DXG and Samsung Galaxy Note	Hello Yes, and one can do that in several ways. One way can be by exporting the culprit pages as images - per example tif non compressed format. Than correct the image in an image editor like Photoshop - a free and very lightweight alternative is Pain.NET (http://www.getpaint.net). Import the edited pages back into finereader, and re-ocr all the images (or if you have ocred all the other pages ocr just the imported ones). Another way can be not to scan directly with finereder (this I think is what you are doing, right?), but to numbered image files with the scanner own application - probably you will get better success rate this way. Than, if one or more pages are not straight, use an image editor to correct them. Import all the images into finereader and ocr them away. Best regards,

11-18-2011, 06:12 AM	#3
roger64 Wizard Posts: 2,608 Karma: 3000161 Join Date: Jan 2009 Device: Kindle PW3 (wifi)	Hi, I'll try the first solution because I have no scanner. I download my PDF image documents from Gallica (Bibliothèque nationale de France). But it would be about time I buy one... Thanks for your tips.

11-18-2011, 08:05 AM	#4
DDHarriman Guru Posts: 860 Karma: 4380 Join Date: Feb 2008 Location: Almada, Portugal Device: Cybook Gen3, Sony PRS 505, Kindle DXG and Samsung Galaxy Note	Hello If you can afford it, try to buy a dedicated book scanner. My advice is the Plustek Opticbook 3600 (http://plustek.com/usa/products/opticbook-series/). Best regards,

11-21-2011, 09:05 AM	#5
DSpider Evangelist Posts: 450 Karma: 343115 Join Date: Nov 2009 Location: Romania Device: PW2 2014	I don't know about FineReader 9 but version 10 has an "Edit Image" button with some very interesting options. You should check it out. But I supposed you could run the images through Scan Tailor, then OCR them. Just make sure they come out grayscale and not in black and white, for more accuracy (because FineReader already has filters in place).

Advert

Advert