Open Access Open Access  Restricted Access Subscription Access

A Review on CAPTCHAs in Use with Focus on Their Existing Vulnerabilities Concerning Web Security

Dayanand *, Wilson Jeberson


Over last decades or so, CAPTCHAs (Completely Automated Public Turing test to tell Computers and Humans Apart)are used to differentiate between humans and bots. In spite of the fact that numerous choices of CAPTCHAs are presented in later a long time but still text-based CAPTCHAs are most predominant among all other choices on the web. CAPTCHA has been broadly utilized over web to guard against assaults from pernicious bots and robotized assaults.Hackers use bots which perform auto registrations on the websites, often resulting in wastage of resources and compromising the security of the website. Therefore, we need a web security component like CAPTCHA which helps us in telling human clients separated from computerprograms. With progresses in Artificial Intelligence and computer vision, it is currently conceivable to break the current CAPTCHAs.InText-based CAPTCHA,more distortion, different orientation and noise make it difficult for humans also. Text-based Captcha can be cracked by using OCR software.It is suffered by relay, random guess, and dictionary attacks. In images-based captcha, sometimes it very difficult to read and are not compatible with users with disabilities. It is also time-consuming to decipher with technical difficulties with certain internet browsers. It may greatly enhance Artificial Intelligence. In audio based CAPTCHA, performance of human users on this task is highly dependent on their fluency. This implies more youthful users whose language abilities are not all around created, non-local speakers or clients with language or learning incapacities may not perform well. In this research work, various reviews of the existing CAPTCHAs in use, with major focus on their existing vulnerabilities has been performed.


CAPTCHA, Web Bots, Web security,robot, motion, artificial problem

Full Text:




  • There are currently no refbacks.

Copyright (c) 2021 Journal of Network Security

  • eISSN: 2395–6739
  • ISSN: 2321–8517