tesseract outputs text, but only nonsense

Question

I am using tesseract on the mac with swedish trained data, http://code.google.com/p/tesseract-ocr/downloads/detail?name=swe.traineddata.gz

I run the following from the command line: sudo tesseract -l swe my.png my.txt

and this is what it outputs in my.txt:

uavum-rn om: mgm.
:mm om names N............
m fw.
<>..,...,.....1,». mm. ^V.m..»...1 W u
|............................ mmm
m«.......
n....... ~.«......«y.= mm
Am...
M-Q-..y...@»~.U.M........»...........   
.;.§............. MYM... WU..
M. www
.<W..L.....w.m.,w»
mm... Hm... ^......... a.....ß..... M
M..
Hm... 3....
>«........
N
1
G
n.......
mmm
mmm »
mmm
MW:-u >«..«.......
M.».....«@>-ms... .a »mm »1
mm... nu .<....-...WMA _..
m........m mm
WW» m
mm w
.-...............u. 
|-...M-11.”.
|........m :>...1.1-1»-.N
Kwwm
M...-«
|.~.»...:-u1.«..... ,-...........
mm M
.-M».....m ...A m...m..<....ß.-.W
.mwwm .M M»-..U..........k
.....-W... .W-;-1

Is there some parameter I miss, one I am doing wrong?

Thanks.

Answer 1

It's been a while since I played with this engine but your note rings a bell with me so I checked the site. I think you need to use this file Swedish language data for Tesseract 3.02 because I suspect you're using the training data from the previous release.

If I got the root of your problem, mark my answer will you? ;)