 |

11-16-2009, 09:02 AM
|
 |
Grand Master
|
|
Join Date: May 2005
Location: USA, Oregon
Posts: 3,018
|
|
How did LeoSleep work for RSC?
Anyone know how LeoSleep worked for runescape classic?
Did it detect what letter were on the screen for the captcha?
Or did it read the memory of the client?
__________________
I DO NOT MM, I DO NOT BUY STUFF, I DO NOT SELL STUFF, I DO NOT PLAY RUNESCAPE!
|

11-16-2009, 02:03 PM
|
|
Legend
|
|
Join Date: Jan 2007
Posts: 10,902
|
|
Re: How did LeoSleep work for RSC?
Didn't you say somewhere u had a copy of it?
__________________
PM me with any issues.
I do not MM or trade anything.
Anyone claiming to be me in a trade or MM is an imposter.
|

11-16-2009, 11:21 PM
|
 |
Grand Master
|
|
Join Date: May 2005
Location: USA, Oregon
Posts: 3,018
|
|
Re: How did LeoSleep work for RSC?
LeoSleep 1.0.3
Quote:
.____ TheIntensity _________.__ Proudly Presents, v1.0.3
| | ____ ____ / _____/| | ____ ____ ______
| | _/ __ \ / _ \ \_____ \ | | _/ __ \_/ __ \ \____ \
| |___\ ___/( <_> )/ \| |__\ ___/\ ___/ | |_> >
|_______ \\___ >\____//_______ /|____/ \___ >\___ >| __/
\/ \/ A RuneScape \/ Classic \/ OCR \/ |__|
http://www.theintensity.com/
Intensity's Power!
== Version History ==
v0.1 - Private release
v0.2.1 - First public release
- minor bug fixed
- multiple folders added
v0.5.0 - Improved front gui
- Added folders section
v0.8.2 - Client mode added
- Main timer code rewritten for performance
- New splash screen image =)
v0.8.5 to v0.8.6
- Bug fixes
v0.9.2 - Intervals added
- Gui tweaks
- Added option to save a copy of hc.bmp
v0.9.5 to v0.9.7
- Small improvements and tweaks
v0.9.8 - Added option to split connected letters
- Added batch files for low priority (windows nt based
systems only only, most likely :P)
- Hopefully either Leosleep or LeoSleep-comp will work
fine on windows 9x systems
v1.0.0
- Fixed some shtuff =)
- Should be final release for a while
v1.0.1
- Hopefully fixed the crashing problem =D
v1.0.2
- Added a button to show or hide the settings
- Added some options regarding the system tray and errors
- Added transparency slider, won't work on Windows 9x
v1.0.3
- Fixed bug that took too long to write the sleepword
== Simple guide ==
NB: May not work on Windows 9x based systems.
= Modes =
There are 3 modes this can be run in, client mode, bot mode,
or off.
- Off
This is rather self explanatory, I think :P
- Client Mode
Use this only when using the legit runescape classic client
from runescape.com, most bots won't work with this mode, as
they are designed to use bot mode. Some bots will work, and
those that are likely to will move the screen up when you
sleep.
To use this mode, simply check the Client check box, then
drag the crosshair over runescape, like you would for using
scar.
This mode is generally only used for scar + legit/just clients.
- Bot Mode
Use this mode when your bot saves a file called HC.BMP when
you sleep. Simply select the Bot check box, and add the
folders where your bots save the hc.bmp. Usually this is where
run.bat or where the bot's .jar file is located.
This mode is used for ar/ix, sbot, sts, tlm, ab, abot,
gbot, rr, etc.
= Options =
- "Attempt to split joined letters"
LeoSleep will try to split joined letters. Although not
necessary for anything, it will improve the accuracy beyond
that of FOCR's normal accuracy.
!! NOTE: May slow down your computer. !!
- "Save copy of hc.bmp"
This option saves a copy of the sleepword in a folder called
hcs where your LeoSleep.exe is. The format is <word number>.
This option isn't really needed for autoing, but fun to have.
Compare LeoSleep's accuracy with your other OCR's.
!! NOTE: May slow down your computer. !!
- "On Minimise"
When LeoSleep is minimised it will do one of three things,
depending on this option. If systray is selected, LeoSleep
will minimise to the system tray down near the clock. If
normal is selected, it will minimise like any other
application, and if hidden is selected it will hide itself
from the users view.
!! NOTE: When you use hidden, hold Control + Shift + Alt + F2
To un-hide LeoSleep. !!!
- "On Error"
This option will decide how LeoSleep acts when an error occurs.
When on Window, LeoSleep will use it's own window to display
errors and attempt to continue operations in the background, but
this may cause a few windows to open. MsgBox will display a simple
dialog displaying the error, and will do nothing else until the
user has pressed Ok. Ignore will try to ignore any errors.
== FAQ ==
Q: It says I am missing file x.
A: Download the needed files zip from the leosleep page, or
search google for the file.
Q: How do I report errors, bugs, or features I would like?
A: Come to our forums (theintensity.com) and post a topic in
the leosleep section with a detailed report of what is the
problem.
Q: I would like Leo to have my children.
A: ...
== Info ==
By Leo,
http://www.theintensity.com/
"Intensity's Power!"
|
Also has word dictionary.
http://uppit.com/v/VUE7RP6W
__________________
I DO NOT MM, I DO NOT BUY STUFF, I DO NOT SELL STUFF, I DO NOT PLAY RUNESCAPE!
|

11-17-2009, 09:37 AM
|
|
Apprentice
|
|
Join Date: Aug 2005
Posts: 806
|
|
Re: How did LeoSleep work for RSC?
The DLL has 3 function;
Code:
getString
loadDictionary
loadModel
The 'model' file looks like coordinates, that plus the getString function and the dll file is called OCRdll.dll. I'd have a stab at "Reads the characters from the screen"
|

11-17-2009, 03:15 PM
|
|
Newcomer
BANNED
|
|
Join Date: Nov 2009
Posts: 14
|
|
Re: How did LeoSleep work for RSC?
It's titled "A Runescape Classic OCR". And I'm sure you know what OCR is/does. So that's that.
|

11-22-2009, 10:14 PM
|
|
Member
|
|
Join Date: Dec 2008
Posts: 91
|
|
Re: How did LeoSleep work for RSC?
wow... this brings back memories for sure :-( anyone else remember theintensity.com? anyway, it works via a support vector machine, very similar to my own ocr. if you don't know what that is, it's using a linear classifier to classify input data ( a feature vector) which was obtained from a feature extractor on the image and returning which class it best fit into. svm's are used in ai because classifying is a key part to machine learning :-)
if you're wondering...
one class for each letter.
model.txt contains the training data for the classes btw
the only mystery left is what features it extracted? edges? corners?
|
 |
|