Captcha Guesser is a PoC to explore AWS Rekognition capabilities. Main goal is to solve some simple captchas generated by different providers.
Before sending images to AWS Rekognition, they are preprocessed for obtaining better results.
THIS PROJECT IS NOT BEING AND SHOULD NOT BE USED EXCEPT OF PERSONAL, NON COMMERCIAL USE. USE AT YOUR OWN RISK.
-
~70-80% success rate
, mostly issues with confusingJ<>I
andD<>O
-
NOT WORKING
- almost 0 success rate -
NOT WORKING
- almost 0 success rate
- NodeJS >= v8.x
npm install
# Check .env.sample
touch .env
echo 'accessKeyId=KEY_ID' >> .env
echo 'secretAccessKey=SECRET_KEY' >> .env
echo 'region=eu-west-1' >> .env
https://docs.aws.amazon.com/sdk-for-java/v1/developer-guide/credentials.html
TBD (check bin/cli.js
for now)
# Test Moldcell with predefined assers
bin/cli.js test -p moldcell
# Test Moldcell with live data (obtained 100 times)
bin/cli.js test -p moldcell --live -i 100
- Improve orange
- Improve unite
- Add orange live testcase
- Add unite live testcase
This might never happen ;)