Text Based CAPTCHA the Guidelines to Enhance the Robustness of CAPTCHA using Chosen Plain Text Attack Method

In general, websites face a variety of security issues. To address these challenges, web pages are protected by specific security measures. One such measure is CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart), a technique used to identify human users and allow them to proceed within the selected webpage. There is ongoing research aimed at distinguishing between human users and machines. To deceive machine learning models used in CAPTCHA systems, a common observation is: “The higher the cracking rate, the poorer the quality of the CAPTCHA.” These results contribute to refining the design guidelines of CAPTCHA systems. Cracking CAPTCHAs using chosen plaintext attacks and analyzing the experimental results can help in designing more robust CAPTCHA mechanisms.