How to bypass a (simple) Captcha with Python3 and Pytesseract?

A CAPTCHA, which stands for Completely Automated Public Turing test to tell Computers and Humans Apart, is a test used by websites around the world to determine if a user is truly human.

According to this excellent article from Cloudflare posted in May 2021, with 4.6 billion Internet users every day, and a captcha solved on average once every 10 days, that's nearly 500 million captchas solved every day.

If it is possible to solve these challenges by using third party services, such as the company 2captcha, it quickly becomes slow and expensive, since it takes an average of 30 seconds per resolution, and 2 USD for 1000 captchas solved. Not to mention the ethical implications.

In this article, we will see how to bypass a (simple) CAPTCHA in a completely programmatic way, with Python and the Pytesseract library.

The code is available in full here: https://gist.github.com/lobstrio/da95d31bff3f83a5e95ee9daeb253107#file-bypass_simple_captcha_pytesseract-py

Prerequisites

In order to complete this tutorial from start to finish, be sure to have the following items installed on your computer.

1. python3
2. SublimeText

You can click on the links below, which will take you either to an installation tutorial or to the site in question.

To clarify the purpose of each of the above: python3 is the computer language with which we will scrape the pdf, and SublimeText is a text editor. Sublime.

Let's play!

Setup

We will proceed as follows:

1. Install open-cv
2. Install pytesseract
3. Install tesseract
4. Download the captcha

For the first 2 libraries, you just have to type the following commands in the console:

f
$ pip3 install opencv-python
$ pip3 install pytesseract

Finally, you need to install tesseract, which is an OCR, the acronym for Optical Character Recognition, i.e. the technology that will allow you to decipher the characters of the Captcha.

Mac OS

f
$ brew install tesseract

Linux

f
sudo apt update
sudo apt install tesseract-ocr
sudo apt install libtesseract-dev

And all the detailed installation information here: https://stackoverflow.com/questions/50655738/how-do-i-resolve-a-tesseractnotfounderror

And here we are, the libraries are installed!

Finally, we are going to download our (simple) captcha. Go to this captcha generation site: https://fakecaptcha.com/. Type the keyword of your choice. Finally, download the image, to keep preciously:

And there you have it, the bookstores are set up!

Let us now decipher this image. And prove our humanity. Without any human intervention.

🤖

Step-by-step Guide

We will go through 3 distinct steps:

1. Resize
2. Close
3. Threshold

With the 3 transformations as follows:

1. Resize

First, we will resize the image. Resizing the image allows the OCR algorithm to detect the character or number strokes in the input image.

The code as follows:

f
filename = 'lobstr.jpeg'
img = cv2.imread(filename)
gry = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
(h, w) = gry.shape[:2]
gry = cv2.resize(gry, (w*2, h*2))

2. Close

Closing is a morphological operation to remove small holes in the input image. If we look carefully the characters 'l' and 'b' are composed of many small holes.

Code:

f
cls = cv2.morphologyEx(gry, cv2.MORPH_CLOSE, None)

3. Threshold

We will apply a simple threshold to binarize the image. Our goal is to remove any remaining artifacts from the image that impair readability.

Code:

f
thr = cv2.threshold(cls, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]

4. Decrypt

Finally, once the image is clear enough, we decode the message with the tesseract library for Python.

And we print the decoded message in the console:

f
txt = image_to_string(thr)
print(txt)

Code

Here is the complete code:

f
import cv2
from pytesseract import image_to_string

# pip3 install opencv-python
# pip3 install pytesseract
# brew install tesseract

filename = 'lobstr.jpeg'
img = cv2.imread(filename)
gry = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
(h, w) = gry.shape[:2]
gry = cv2.resize(gry, (w*2, h*2))
cls = cv2.morphologyEx(gry, cv2.MORPH_CLOSE, None)
thr = cv2.threshold(cls, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
txt = image_to_string(thr)
print(txt)

To execute the code:

1. Download the .py code
2. Change the path of the image
3. Launch the script via the command line

And this is what will appear directly on your terminal:

f
$ python3 bypass-captcha-pytesseract-tutorial.py
lobstr

Eureka!

✨

Limitations

This code will allow you to quickly convert images, generated from https://fakecaptcha.com/, into text. And thus to bypass the possible challenges posed when scraping data.

However, this script only works for simple captchas. Let's try it with an Amazon Captcha, below:

We run the script:

f
$ python3 bypass-captcha-pytesseract-tutorial.py
nrrth

Which simply does not work.

🤷‍♀️

Conclusion

And that's the end of the tutorial!

In this tutorial, we've seen how to bypass a simple Captcha with Python3 and Pytesseract, programmatically.

If you have any questions, or if you need a custom, robust and scalable scraping service that can bypass even the most robust captchas, contact us here.

Happy scraping!

🦀