Solving CAPTCHAs using Convolutional Neural Network
This project aims at predicting and solving python-generated CAPTCHAs, using Convolutional Neural Network to detect the vulnerability of the CAPTCHAs. By running the program, a specific amount of CAPTCHAs will be generated in a folder which then is used as the training set of the network.
To run the code, the following python packages are required:
- captcha
- opencv-python
- opencv-python-headless
- opencv-contrib-python-headless
- keras
Also, the following Linux packages are required:
- libsm6
- libxext6
To run the code open [Captcha_Solver_Alphanumerical.ipynb] or [Captcha_Solver_numerical.ipynb] in Jupiter notebook and then run them.
- This work is based on Deep-CAPTCHA: a deep learning based CAPTCHA solver for vulnerability assessment
- For more information please contact: zaraanry@gmail.com (Zahra Nouri) or m.rezaei@leeds.ac.uk (Mahdi Rezaei)