No Captcha

This repo serves to help with webscraping websites with simple captchas, you can easily create datasets, train models and make predictions

Creating Dataset

First you will need labeled images, with the respective letter they are. The way the code deals with this is with a folder with folders named from A to Z inside, with the letter images inside these folders.

More About the code

https://www.kaggle.com/code/iuryck/reading-captchas-with-convnets

Requirements

scikit-learn==0.24.2
keras==2.6.0
opencv-python==4.6.0.66
imutils==0.5.4