Tesseract not recognizing a single digit in .bmp files

[GoogleCodeExporter]

What steps will reproduce the problem?
1.Both files attached are very similar and contain each '8'.
Called by "tesseract im1/2.bmp test -psm 10" (10 for 'single character')
Tesseract finds '8' in im1.bmp and ' ' (blank, 0x20) in im2.bmp.
Option 'digits' has no effect.

2.
3.

What is the expected output? What do you see instead?
expected: '8' for im2, delivered: ' ' (blank, 0x020)

What version of the product are you using? On what operating system?
tesseract 3.02 portable, Windows 8.1, 

Please provide any additional information below.

Additional info: using psm-option 6 (6 for 'single uniform block of text')
Tesseract recognizes the '8' in im2.bmp.

Original issue reported on code.google.com by Ree...@web.de on 3 Feb 2015 at 9:54

Attachments:

• im2.bmp

[GoogleCodeExporter]

Original comment by Ree...@web.de on 3 Feb 2015 at 10:07

Attachments:

• im1.bmp

[GoogleCodeExporter]

Works with current code (in repository)

Original comment by zde...@gmail.com on 12 Apr 2015 at 7:45

• Changed state: WorksForMe

[GoogleCodeExporter]

would you please name the 'current code'-filename/version?

Original comment by Ree...@web.de on 15 Apr 2015 at 9:04

[GoogleCodeExporter]

I am not sure if I understand your request: current code it in git repository:
https://code.google.com/p/tesseract-ocr/source/checkout

Original comment by zde...@gmail.com on 15 Apr 2015 at 9:10

[GoogleCodeExporter]

I have used tesseract executable/compiled Version 3.02 portable for Windows 8.1.
Does a newer executable/compiled Version exist which works with psm option 10? 

Original comment by Ree...@web.de on 15 Apr 2015 at 11:27