ShaneNolan / MarkCAPTCHA

An Examination on the Robustness of Alphanumeric CAPTCHA as a Challenge Response Test using Machine Learning.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

MarkCAPTCHA

An Examination on the Robustness of Alphanumeric CAPTCHAS using Machine Learning.

Read Thesis (PDF).

Arguments

Arguments supplied when running MARKCAPTCHA.

required arguments:
  --config              the filename of the json config file.

optional arguments:
  -h, --help            show the help message.

  --imageprocessing     preprocess the supplied config images.

  --build               build a CNN model using the supplied images.

  --predict             a CAPTCHA image in base64 format.

  --show                display the image processing and prediction output. [Incompatible with Docker]

Build

Docker Container

Build MARKCAPTCHA container.

docker build --tag markcaptcha-container .

Run the container.

docker run markcaptcha-container [ARGS]

Run without Docker

python markcaptcha.py [ARGS]

Usage Examples:

--config simplecaptcha.json --predict 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

--config captcha03.json --predict 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

--config pastebin.json --predict /9j/4AAQSkZJRgABAQEAYABgAAD//gA+Q1JFQVRPUjogZ2QtanBlZyB2MS4wICh1c2luZyBJSkcgSlBFRyB2NjIpLCBkZWZhdWx0IHF1YWxpdHkK/9sAQwAIBgYHBgUIBwcHCQkICgwUDQwLCwwZEhMPFB0aHx4dGhwcICQuJyAiLCMcHCg3KSwwMTQ0NB8nOT04MjwuMzQy/9sAQwEJCQkMCwwYDQ0YMiEcITIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIy/8AAEQgAIwB4AwEiAAIRAQMRAf/EAB8AAAEFAQEBAQEBAAAAAAAAAAABAgMEBQYHCAkKC//EALUQAAIBAwMCBAMFBQQEAAABfQECAwAEEQUSITFBBhNRYQcicRQygZGhCCNCscEVUtHwJDNicoIJChYXGBkaJSYnKCkqNDU2Nzg5OkNERUZHSElKU1RVVldYWVpjZGVmZ2hpanN0dXZ3eHl6g4SFhoeIiYqSk5SVlpeYmZqio6Slpqeoqaqys7S1tre4ubrCw8TFxsfIycrS09TV1tfY2drh4uPk5ebn6Onq8fLz9PX29/j5+v/EAB8BAAMBAQEBAQEBAQEAAAAAAAABAgMEBQYHCAkKC//EALURAAIBAgQEAwQHBQQEAAECdwABAgMRBAUhMQYSQVEHYXETIjKBCBRCkaGxwQkjM1LwFWJy0QoWJDThJfEXGBkaJicoKSo1Njc4OTpDREVGR0hJSlNUVVZXWFlaY2RlZmdoaWpzdHV2d3h5eoKDhIWGh4iJipKTlJWWl5iZmqKjpKWmp6ipqrKztLW2t7i5usLDxMXGx8jJytLT1NXW19jZ2uLj5OXm5+jp6vLz9PX29/j5+v/aAAwDAQACEQMRAD8A90mjvUlaW2mSRT1gmG0Dt8rqMj1OQ2egxUEuqrbxN9rjktXXks65iOP+mg+UA9BuwenHIrSJwCcZ9hXM+K/G2m+FLPfch5LuRf8AR7VR80rf0Gep/LNJtbMVtdDoZfMeMNA67gdwB+649Ce31H69CsMyzISAVYHDI3VT6H/PuOK8e8DeONW1fXprCSyFnFBG8jfZIpHEQDABWhJII3EA7QG6nI5NegavdapaaXc3lnp9vqokgZysTgRzLsGCVJPJ5xtLZAwe2GpJoNbnRLNG8skSt88eNykYIB6H3HXn2I7GsLQPGmi+JdRvbLTJ3kltMF9yFQ4zjK55IB4PA6ivFdD+I9yupmfVbN/ssbhkfT8x/YgSAdicptPQgjJyec0ajdv4U8WWfjHRb6LUrC7kJkkiURgsR+8jdR9xiOQCPftWvsuhpy6H0VVffNB/rR5sf99FO4fVR1+o9egFZNn4mjurW0uPsjtDchWWaCRJIwGGQTkhhgdcqOeOeMtm8SZuI7dI47Z5QNn2hwzkk4G2NT8wJGPvDnPHHODqRi7NkLU3o5ElQPG6uh6MpyDWbeT6bdSeWQ1zcREqPs2TJE3Qjev+rzjHJUHBzwDXnnxK17VvCsNjcwqtyly7K63g3I2BxlF2qev8S54HpVI+PPE2heItH0/U30y503UVj8trFNoRGO35Tn+E9R7YBHWl7TrEHKC8z02KDVfLk8qZooyMRx3RWSQe25RhfqxkzwTzkGzp0dxH5oulm87jdI8wdH642gYxgdfkXPv1qvaXEkoeR7vETHECRpl2UEjfzuJDHJGMjaFPHNFxexQSpHPK4ZhlBLMIuPX5eg/3sH0BPFL2jfTX+u3+Rd7rsjWorn9e8TJovh261Db5ksaZj/dP5bMThRuxgjkc55rN8E+PF8WzT2z2P2a4hjDkiTcrDODjgEdvWtCLp7HZUVmw+IdHnvZLKPU7U3Ub7Gh80Bt3oAev4UUAXXldWIFvI4/vKVwfzNeW3mjeING8R6p4kW00jVC7PJ/pjt5sEYydqfMQABx07V6syq6lWUMp7EZqK4tYLuzltJ41e3mjMTxnoykYI/Kpab2ItK+jPCvhLNeHVtSuhY3Mgu/kN3BwIm5cg8Hrx2Nenz6O8tw0sN1cRAHMkEaMsczZyTKoTDEg8kjngVp6T4bsvD1s9rokMNrbvIZWR1eQ7yACcls9AOKmudOvLmQP/aAiIG0+VDtJH13Z/WkqKbu5WKld7P8AA8K8WadeeCvGketw6cqWFw5DJszC+R86YwuAQTwQOc44Gat6z8PLa802PUtC1OGP7XGJjaPuRJASSNvGBjgYJIyDyK9S8RaBZ6lpkllrMy/ZZGBV2MjHcORtzISG4I4HIJHfFM02wvfD1l5emWkNvpsYYJC+RknkSHJLKOgOTxkkrxXSo2Sae34/8H8QblZJNXOb8AfaPC/h57O9vf3gkLqm35EJHCb9pBzgnqO9dgNXur6GaNYrUxmNlJuMBHP90DcdwOevQ+tV4Yl+1JBfQRz35X/V3EQCyHbkiMlgpAHUqvoSM1ca2HBjtXiZRhCkkuEHoqmMqB6ccdsVySdrtPX1t910i1TlLd/qeT/Fi1MPh2xZlt42+2H91bgBBlDyAvy9scgtwMkjFc7r0F5oup6Breo3qanBd2kTpIUAaDCjK4BHzKCME9fw49l8QWNjqenxWmqpb38UUnmC3knYyb+cH5VUjALcH+lVf7DtLuApf6LaXllDGqWsZmISNEXsDkhjknnB4z1zUycm7v8AQh4eSW4unTPHpsc/lwXtjIodPs8ofCBRhlSTHzHn+HtxgEV0em6zohItLWaK3kyo8mRTExYjgYbG49Ome1ZenXaaXp8NtAkNnEg/dxRyb4+ufugOVHTjI9u9aUn2HVbZP7Rls96jMckUxVlz6ZwV7cd+9bQqRfxP8V+Qm1Hdo474xaiy6bp+lREl7iUysqjJIXgD8S36Vy95HqPg/wAYSadosJN7d6fDapt67yqbmHbOUPPbOa7658M6ZeavZSzC3vzC6CJ/OxIihs8heGHsVJPJLAcV1TaZZJci9isLX7ZGhWOXy1DDjpuxkD/Gm0ujuClfVHiPhjw6P+FmwacZBcCxk8y4fqC6DLD6b+KK7r4eeE9U0XU9V1DWY1W5uCFRldW3Aksx4Pc4646UUFHoFFFFABVXUpnt9KvJom2yRwO6nGcEKSKKKqHxIUtmQ6RCj2Nteuu+6ngVnlY5Y7gCQCei552jAHpVmICO4kiXIQIrBc5wSW6eg4HFFFVN+8wh8JXht4bhLyznhjltklCrFIoZQNqNjB7Anj04x0FR3um2a6fIwt0LRIXVm5YkA4LE8t75znvmiis62zfl+gRim1dFiPS7CKMIlpDtHTKAn8zUgsbQdLWAf9sxRRU8kewWXNzdQ+x24+7EsZ9Y/kP5jFPMETw+TIgkjPVZPmz35zRRQoRWyK5n3IlsLWNw0UIiIOf3RKA/ULjP41ZooppJbA5N7sKKKKYj/9k=

Testing

Arguments supplied when running an accuracy test.

required arguments:
  --config              the filename of the json config file.
  --folder              path to image testing folder.

optional arguments:
  -h, --help            show the help message.

  --sample              sample number of images to test.

  --show                display the image processing and prediction output.

  --print               display the accuracy output.

Run accuracy test script.

python tests/test_Accuracy.py [ARGS]

Usage Examples:

--config captcha03.json --folder C:\MarkCaptcha\markcaptcha\data\captchas\captcha_03\captchas\testing

--config pastebin.json --folder C:\MarkCaptcha\markcaptcha\data\captchas\pastebin_captcha\captchas\testing

--config simplecaptcha.json --folder C:\MarkCaptcha\markcaptcha\data\captchas\really_simple_captcha\modified\tmp\testing

About

An Examination on the Robustness of Alphanumeric CAPTCHA as a Challenge Response Test using Machine Learning.


Languages

Language:Python 70.4%Language:PHP 29.3%Language:Dockerfile 0.3%