Skip to content

A python application to solve the SIS Portal CAPTCHA using a custom trained model with TensorFlow and Keras.

Notifications You must be signed in to change notification settings

nmnjn/Captcha-Solver

Repository files navigation

Captcha-Solver

A python application to solve the SIS Portal CAPTCHA using a custom trained model with TensorFlow and Keras.


Tools

  • Python 3
  • OpenCV
  • Keras
  • TensorFlow

Installation

Download the Source Code
git clone https://github.com/naman17/Captcha-Solver.git
(or download the zip)
Install the Requirements
cd Captcha-Solver
pip install -r requirements.txt
(Optional) Download a set of CAPTCHAs from the SIS Portal:
make download

Running the Script

Step 1 - Extracting Dataset:
make extract-data
Step 2 - Training the Model:
make train-model
(Optional) To test the Script with a URL:
make captcha-test
(Optional) To start a server:
make server

The server will start on localhost:5000. To test the server send a get request on /solve?url=<append-url-here>


As a proof of concept you can test the script on http://139.59.7.182:8001/solve?url= by appending the CAPTCHA Image URL from the SIS Portal.


(Optional) To clear all the datasets and model generated :
make clean

Running with Docker

docker build -t sis-captcha-solver .
docker run --rm -it -p 5000:5000 sis-captcha-solver

The server will start on localhost:5000. To test the server send a get request on /solve?url=<append-url-here>


About

A python application to solve the SIS Portal CAPTCHA using a custom trained model with TensorFlow and Keras.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published