Skip to content

Instantly share code, notes, and snippets.

@christianroman
Created May 30, 2013 16:02
Show Gist options
  • Save christianroman/5679049 to your computer and use it in GitHub Desktop.
Save christianroman/5679049 to your computer and use it in GitHub Desktop.
Bypass Captcha using 10 lines of code with Python, OpenCV & Tesseract OCR engine
import cv2.cv as cv
import tesseract
gray = cv.LoadImage('captcha.jpeg', cv.CV_LOAD_IMAGE_GRAYSCALE)
cv.Threshold(gray, gray, 231, 255, cv.CV_THRESH_BINARY)
api = tesseract.TessBaseAPI()
api.Init(".","eng",tesseract.OEM_DEFAULT)
api.SetVariable("tessedit_char_whitelist", "0123456789abcdefghijklmnopqrstuvwxyz")
api.SetPageSegMode(tesseract.PSM_SINGLE_WORD)
tesseract.SetCvImage(gray,api)
print api.GetUTF8Text()
@anhtuan0030
Copy link

@forum2k9 I solved the first one with convert and Tesseract:

$ convert ofdbmf.jpg -colorspace Gray -blur 0 -level 0,60% ofdbmf-1.jpg
$ tesseract -psm 8 ofdbmf-1.jpg -
OFDBMF

I posted the details at https://mathieularose.com/decoding-captchas

Olá, alguma sugestão de texto cruzado por várias linhas curvas com o mesmo peso?
download-3
download-4
download-6
download-7
download-2
download-5

Have you made any progress? I have this same captcha.

Hello sir, do you resolve this? can you share with us pls?

@forum2k9 can you help us too?

Thanks!

@neodouglas
Copy link

how i can solve that?
007444
002952
002963
004058
004967
005589

@AlexanderMortimer
Copy link

it is not working for my captcha @jtanori
1

@DiMiTriFrog
Copy link

it is not working for my captcha @jtanori
1

Any solutions nowadays?

@NotTrueFalse
Copy link

2jB4
6mKp
ByLt
HZj0
kumd
mVVZ
NJJ6
NW3K
tEvz
x1CP

Hi,
I want solve a Captcha similar yours.
Did you solve it?
i will appreciate if you help me .

@AlexanderMortimer
Copy link

AlexanderMortimer commented Jan 27, 2024 via email

@NotTrueFalse
Copy link

I was joking, I'm actually making a model to solve them, don't worry

@AlexanderMortimer
Copy link

AlexanderMortimer commented Jan 27, 2024 via email

@NotTrueFalse
Copy link

It's done, but it work only with the type of captcha I posted here, If you want me to make a model for your type of captcha, I need a dataset with at least 10k captcha.
If you don't have it, send me an image, I'll find how to gen a lot of captcha, make a model then tell you how to use it.
contact info : enderty on discord

@NotTrueFalse
Copy link

NotTrueFalse commented Jan 28, 2024

Sounds good. Please let me now whener it completed

https://github.com/NotTrueFalse/Captcha_solving/

@ytrezq
Copy link

ytrezq commented Feb 1, 2024

It's done, but it work only with the type of captcha I posted here, If you want me to make a model for your type of captcha, I need a dataset with at least 10k captcha. If you don't have it, send me an image, I'll find how to gen a lot of captcha, make a model then tell you how to use it. contact info : enderty on discord

@NotTrueFalse hey : those kind of captchas are automatically generated by https://captcha.com/java-captcha.html. You could get millions of such annotated captchas.

@NotTrueFalse
Copy link

Thank you, I'll make another branch so you can generate another type or captcha, I'll have to update the pre-processing part though but it should be fine.

@ytrezq
Copy link

ytrezq commented Feb 1, 2024

@NotTrueFalse my intent wasn t generating the dataset myself.

@NotTrueFalse
Copy link

You want me to do all the step to create the model? At this point I could just create a website and make a cheap api for people

@ytrezq
Copy link

ytrezq commented Feb 1, 2024

@NotTrueFalse I lack the knowledge to train an ai and I m bad at advanced databases scenarios.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment