Correct recognition with one line but not multiple lines

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Take a screen shot from www.raspberrypi.org and past it in Gt-text.
2. Set Tools | Area Text OCR
3. Click and drag 1 line or multiple lines of text

What is the expected output? What do you see instead?

Correct text.  I see correct text for 1 line, but if I mark multiple lines is 
see mostly gibberish, very few characters recognized.

What version of the product are you using? On what operating system?
1.4.4 on Windows 7

Please provide any additional information below.

See attachments.

Original issue reported on code.google.com by myko...@gmail.com on 9 Sep 2012 at 10:31

Attachments:

• gt-text-attach1.png
• gt-text-attach2.png

[GoogleCodeExporter]

Original comment by dto...@gmail.com on 9 Sep 2012 at 8:47

• Changed state: Accepted

[GoogleCodeExporter]

Original comment by dto...@gmail.com on 5 Jan 2013 at 2:17

• Changed state: Started

[GoogleCodeExporter]

Visually, the lines are cut, but copied in clipboard without this line break.

Original comment by dto...@gmail.com on 22 Feb 2013 at 11:59

• Changed state: Done