Introduction :

CAPTCHAs (Completely Automated Public Turing Tests to Tell Computers and Humans Apart) are something that almost every internet user has encountered. Many users are presented with strange-looking, stretched, fuzzy, coloured, and shape distorted visuals that look more like a Dali painting than English text while signing in or creating an account, making an online purchase, or even publishing a comment.

With the growth of Artificial Intelligence, Machine Learning and Computer Vision, the need for strong captchas has increased, so that Computer Vision models are unable to automatically detect these captchas.

So for the same we bring to you the hand written captcha detector which has a pretrained model consisting of 19 letters:

'A','D','E','G','H','J','K','M','N','P','R','S','W','X', 'Z', 'a', 'b', 'd', 'g'

and 7 emojis mapped as:

Checkmark : 1	Cloud: 2	Croissant: 3	Heart: 4	Laugh: 5	Smile: 6	Sun: 7

Setting up the project :

1. Creating virtual environment:

python -m venv venv

2. Activate virtual environment

Linux:

source venv/bin/activate

Windows:

./venv/Scripts/activate

3. Install dependencies

pip install -r requirements.txt

4. Install the package:

pip install -e .

5. Move to examples folder:

cd examples

6. Run main.py :

python main.py

How to use :

After setting up project just place your images in line folder under the data directory and run the main.py file again. The program will show you images and its output in terminal one by one. If you think that your images is not recognised clearly by program just change the threshold value according to your needs for best output.

NOTE: The model is trained for the above mentioned letters so give them only in input images for best output.

Made and maintained by :

Ankur Agrawal	Aryaman Gupta	Vivek Agarwal