NCTU new_e3 captcha_verify

scrape.py is a Web Crawler program to download huge training captcha file.
tt3.py is a background noise reduction program.
tt44.py is a digit location detection program.
tt55.py is a size normailizing program.
train.py is a traning CNN model program.
predict.py is a predict captcha program.
selepublic.py is a final intergrating login program.
selepublic_v2.py is multi-thrading implemented version.
split_digits_in_img.py is a neccessary code file for selepublic.py and selepublic_v2.py make sure you are downloaded.

1. In selepublic.py you should input your own username and password in advanced(#215,216), then selepublic.py will input them in execution step.
   
2. You only need to download selepublic.py and cnn_model.h5. ##NOTICE you should put those two files in same directory make it can load model.
   
3. NOTICE you should install the chrome driver of selenium.
   
4. You should check the path setting in the code and make the properly setting.
   

Demo video https://youtu.be/KA4EvdrpzaM