Pictograms and Python

[Lad]

Hi,
did anyone try to convert a Pictogram ( = a picture with text) to text?
Lad

 

[Igorx Ivanov]

Did anyone try to convert a Pictogram ( = a picture with text) to text?

Do you mean something like "anytopnm <Pictogram-file> |
pnmscalefixed -width ${COLUMNS} | ppmtopgm | pgmtopbm | pbmtoascii"?

 

[Diez B. Roggisch]

Hi,
did anyone try to convert a Pictogram ( = a picture with text) to text?
Lad

From this and your last post I assume that you want to perform some
registration stuff that is protected against scripts like yours using
images with distorted characters that a human beeing can decipher and
enter, but that are hard to identify using OCR techniques. They're called a
CAPTCHA.

As these protections are there for a good reason and I personally loathe
spam, I hope you neither succeed nor get help from this community.

In the (I believe unlikely) case of your plans to automate access to that
site are of good intentions, I'm sure you can make a deal with the people
running that service so that you can access it through some RPC mechanism.

 

[Diez B. Roggisch]

Of course, these only matter if the number of documents is large. I don't

know what is the likelihood that the OP wanted the help for nefarious
purposes, but I found it presumptuous of you to assume so just off-the-bat
as you did. I would feel rather insulted had I been the OP. I'm sure you
can understand.

If I may quote the OP's first post on performing http requests from two days
ago:

"""
I have a sript that downloads a webpage. According to the picture  on 
this webpage I need to pass a parameter to this , running script a few
lines later.
"""

When I first read that, I already assumed that he wanted to circumvent some
registration CAPTCHA - but I wasn't sure, so I didn't say anything. But the
question for extracting characters from an image makes it clear IMHO. Of
course I might be wrong, and then I'm willing to express my sincere
apologies.

 

[Caleb Hattingh]

Actually there at at least two valid uses for this kind of thing:
1. Digitising hard-copy legacy documents (digitise as in text-editable,
not straight scans)
2. Extract data from graphs in scientific publications for the purposes of
numerical comparison with your own work (one does this anyway, but
normally by hand), and you normally want the axis labels (text) as well
for preserving information about units of measure.

Of course, these only matter if the number of documents is large. I don't
know what is the likelihood that the OP wanted the help for nefarious
purposes, but I found it presumptuous of you to assume so just off-the-bat
as you did. I would feel rather insulted had I been the OP. I'm sure you
can understand.

Keep well
Caleb

 

[caleb.hattingh]

Diez

Ya got me there!

"""
I have a sript that downloads a webpage. According to the picture on
this webpage I need to pass a parameter to this , running script a few
lines later.
"""

Err, ya, I guess I would be suspicious too. Sorry about that!
Keep well
Caleb